Citation: 2009 TCC 563
Date: 20091204
Dockets: 2006-1385(IT)G
2006-1386(IT)G
BETWEEN:
GENERAL ELECTRIC CAPITAL CANADA INC.,
Appellant,
and
HER MAJESTY THE QUEEN,
Respondent.
REASONS FOR JUDGMENT
Hogan J.
SUMMARY OF ISSUE
[1]
GE Capital US (“GECUS”) charged the Appellant a fee for
guaranteeing its debts owing to third-party creditors. The Appellant deducted
that fee in respect of its 1996 to 2000 taxation years. The Minister of
National Revenue (the “Minister”) reassessed the Appellant, denying the
deduction of the fee and adding Part XIII withholding tax because he believed
the Appellant received no economic benefit from the guarantee and, as a result,
the “arm’s length” price for the guarantee would be zero. Part XIII tax is
charged on the basis that the payment of the fee is deemed to be a dividend.
The issue to be determined by this Court is whether the “arm’s length” price
for the service is at least 100 basis points (the Appellant’s position) or zero
(the Minister’s position).
FACTUAL BACKGROUND
[2]
At the commencement of
the trial, the Court received a response to a request to admit the truth of
certain facts (hereinafter the “statement of agreed facts”) as well as the
authenticity of certain supporting documents contained in a joint book of documents.
I shall reproduce in part hereunder the statement of agreed facts before
summarizing the witnesses’ testimony and reviewing the relevant documentary
evidence.
The Appellant and Its Position in the GE Corporate Group
1. The Appellant was originally incorporated as
Genelco Finance Limited under the laws of Canada in 1963. The Appellant changed its name to Canadian General Electric
Credit Limited in 1969 and to Genelcan Limited in 1973. The Appellant adopted
the corporate name General Electric Capital Canada Inc. (“GE Capital Canada”) [Appellant]
in 1988.
2. The Appellant has been an indirect,
wholly-owned subsidiary of General Electric Company (“GE Company”) [GE], a United States company, since incorporation.
3. [GE] carried on industrial businesses and
financial services businesses during the years under appeal through a large
number of legal entities around the world. The corporate structure of [GE] had
separate ownership groups for the industrial and financial services businesses.
4. The financial services ownership group was
organized under General Electric Capital Services, Inc. (“GE Capital Services”)
[GECSUS], an [sic] United States corporation that was a wholly-owned, direct subsidiary of [GE].
5. A principal subsidiary of [GECSUS] was General
Electric Capital Corporation (“GE Capital US”) [GECUS], a United States corporation that was wholly‑owned, direct subsidiary of [GECSUS]
during the years under appeal.
6. During the years under appeal, the Appellant was
a wholly-owned, indirect subsidiary [of GECUS].
7. At all material times, the Appellant and [GECUS]
did not deal with each other at arm’s length within the meaning of the Income
Tax Act (the “Act”).
Business of [GECUS]
8. [GECUS] was a
financial services company during the years under appeal and carried on a
number of financial services businesses, as described as below.
9. In essence, [the GECUS] business model was to
borrow funds from the capital markets at the lowest possible cost and then use
the borrowed funds to lend or lease to other parties on profitable terms.
10. The business of [GECUS] originally related to
financing the distribution and sale of consumer and other products of [GE]. By
the mid‑1990s, however, the types and brands of products financed by [GECUS]
and the financial services offered by [GECUS] had become significantly more
diversified, and very little of the financing provided by [GECUS] involved
products manufactured by [GE].
11. During the years under appeal, [GECUS] provided
a wide variety of financing, asset management, and insurance products and
services in the following five industry segments, directly or through its
subsidiaries around the world . . . :
(i) Consumer Services — private-label
and bank credit card loans, personal loans, time sales and revolving credit and
inventory financing for retail merchants, auto leasing and inventory financing,
mortgage servicing, and consumer savings and insurance services;
(ii) Equipment Management —
leases, loans and asset management services, including sales, for portfolios of
commercial and transportation equipment, including aircraft, trailers, auto
fleets, modular space units, railroad rolling stock, data processing equipment,
containers used on ocean-going vessels, and satellites;
(iii) Specialized Financing —
loans and financing leases for major capital assets including industrial
facilities and equipment and energy-related facilities; commercial and
residential real estate loans and investments; and loans to and investments in
management buy-outs, including those with high leverage, and corporate
recapitalizations;
(iv) Mid-Market Financing — loans and
financing and operating leases for middle-market customers, including
manufacturers, distributors and end users, for a variety of equipment that
includes data processing equipment, medical and diagnostic equipment, and
equipment used in construction, manufacturing, office applications
and telecommunications activities;
(v) Specialty Insurance — financial
guaranty insurance (municipal bonds and structured finance issues); private
mortgage insurance; and creditor insurance covering international customer loan
repayments.
Business of the Appellant
12. The Appellant was a financial services company
during the years under appeal. It carried on in Canada some of the lines of
financial services businesses that [GECUS] carried on in the United States and
other jurisdictions outside of Canada, as described below.
13. In essence, the Appellant’s business model was
to borrow funds from the capital markets at the lowest possible cost and then
use the borrowed funds to lend or lease to other parties on profitable terms.
14. The Appellant was initially formed to assist
in the financing of products manufactured or distributed by General Electric
Canada Inc. (“GE Canada”). By the mid-1990s, however, substantially all of the
products financed by the Appellant and its subsidiaries were products
manufactured by companies other than GE Canada and or its affiliates, and the
Appellant also provided a range of other financial services.
15. At all material times, the Appellant and its
Canadian subsidiaries carried on a number of businesses in Canada. Four of the larger businesses were:
(i) a commercial and transportation equipment
financing business serving manufacturers, distributors and end users with a
broad range of financial products for equipment acquisitions. This business had
locations across Canada and
worked with the transportation, construction, printing, telecommunications,
manufacturing, aeronautics and wholesale/resale distribution industries;
(ii) a fleet vehicle leasing and management
business involved in the leasing and financing of automobiles, light
trucks, vans, buses and medium and heavy duty trucks and trailers to business
users. The Appellant and its subsidiaries also provided specialised
transportation management programmes to clients relating to maintenance
management, national account purchasing, fuel purchasing and insurance;
(iii) a real estate financing business for
a wide range of income-producing properties, including office buildings,
shopping centres, apartment complexes, condominiums, industrial buildings and
warehouses;
(iv) a technology management services business,
which distributed computer products and offered financial arrangements ranging
from sales to day rentals, operating leases and finance type leases. The scope
of products and services included personal computers, mini‑computers and
test and measurement equipment supported by field technicians, networking consultants
and other technology management services.
16. Other businesses carried on by the Appellant
and its subsidiaries during the years under appeal included:
(i) a private label credit card business,
whereby the Appellant or its subsidiary entered into and administered credit
accounts with customers of participating retailers in return for programme
marketing support;
(ii) an auto leasing business providing
retail leasing programmes to dealers of new motor vehicles. The Appellant or
its subsidiary purchased the vehicles and leases from the dealers on a
non-recourse basis;
(iii) short and long term renting and leasing of
over-the-road commercial semi-trailers, chassis and storage containers and
mobile modular office buildings, as well as the sale and financing of such
equipment;
(iv) a full service railcar leasing business
offered to industrial shippers and Canadian railways. The Appellant or its
subsidiary offered a range of management services to lessees including tracking
mileage, arranging for leases, insurance, inspections, maintenance and repair
work, as well as billing, collecting and remitting rents.
17. The Appellant’s business grew rapidly during
the period under appeal . . . .
. . .
19. The Appellant (consolidated) represented
between 2 to 2.3 percent of the consolidated assets of [GECUS] and between
3 and 4.4 percent of the consolidated revenues of [GECUS] during the
period under review . . . .
Debt Securities
20. The Appellant’s business strategy required
substantial amounts of capital, which the Appellant obtained by issuing debt in
the form of commercial paper and unsecured debentures (the “Debt Securities”).
21. The Appellant’s Debt Securities were purchased
exclusively by third parties unrelated to [GE] or [GECUS].
Description of Commercial Paper
22. The Appellant and its predecessor corporations
issued short term promissory notes (also known as “commercial paper”) on
Canadian markets from the 1970s until early 1999.
23. During the years under appeal, the Appellant
issued commercial paper on almost a daily basis. [The Appellant]’s Commercial Paper
Program represented at least ten percent of the Canadian corporate commercial
paper market.
24. During the years under appeal, the Appellant
issued commercial paper under short term promissory note programs established
in 1989 and 1996 (the “Commercial Paper Program”) . . . .
25. The Commercial Paper Program prescribed a
maximum aggregate principal amount outstanding of $7 billion. The actual
aggregate principal amount of promissory notes outstanding during the
Appellant’s 1996 to 1998 taxation years varied from a low of approximately
$1.7 billion to a high of approximately $3 billion . . . .
26. The Commercial Paper Program prescribed a
maximum term of 270 days for the Appellant’s promissory notes.
27. The Appellant issued commercial paper on a
regular basis for a variety of principal amounts, maturities, interest rates
and discounts . . . .
28. The Appellant’s commercial paper was issued in
Canadian and United States
dollars and traded on the Canadian commercial paper market.
29. The Appellant’s issuances of commercial paper
were generally administered by financial institutions in Canada (the “Dealers”).
30. The Appellant made its final issuance of
commercial paper on February 4, 1999.
Description of Unsecured Debentures
31. The Appellant issued unsecured debentures on
European markets from at least 1988 until 1997 (the “Unsecured Debentures”).
32. The Unsecured Debentures typically were issued
for a term of five to ten years and provided for the periodic payment of
interest. The Unsecured Debentures were denominated in a variety of currencies,
including Australian dollars, Canadian dollars, Luxemburg francs, Swiss francs
and United States dollars
. . . .
33. The Appellant issued the Unsecured Debentures
under individual offerings and as part of the multi-issuer Euro Medium Term Note
Program (a simpler and less costly mechanism for issuing Unsecured Debentures
relative to individual offerings). In general, the Unsecured Debentures were
listed for trading on the Luxemburg stock exchange. . . .
34. The Appellant’s issuances of Unsecured Debentures
were fully underwritten by a syndicate of financial institutions (the “Managers”
or “Underwriters”).
35. In respect of each issuance of Unsecured Debentures,
the Managers agreed jointly and severally to procure subscriptions and payment
for the Unsecured Debentures, or failing that, to subscribe and pay for the Unsecured
Debentures on their own account. The Managers, the Appellant and [GECUS]
formalised their agreement in a written Subscription Agreement . . . .
36. Subject to the Subscription Agreement, each Manager
was allotted a specific amount of the Unsecured Debentures as the extent of its
underwriting commitment . . . .
37. One Manager or a small group of Managers was
allotted a significant amount of each issuance as its underwriting commitment
(the “Lead Manager” or “co-Lead Managers”).
. . .
Management of the Debt Securities
39. The Treasury Department of [GECUS] managed the
Appellant’s issuances of Debt Securities. On a daily basis, this department
collected data from the Appellant’s business units and determined the
Appellant’s net cash position for the day.
40. If the Appellant were short on cash for the
day, the Treasury Department approached the commercial paper market through the
Dealers. If the Appellant were long on cash for the day, the Treasury
Department decided how much debt to repay based upon debts that were maturing.
41. The Appellant also funded its acquisitions of
asset portfolios, receivable portfolios and shares in other companies with
commercial paper.
42. In respect of Unsecured Debentures, the
Treasury Department performed an analysis of the Canadian debt position in
total, including maturity levels and the proportions of commercial paper and Unsecured
Debentures. On the basis of this analysis, the Treasury Department looked for
opportunities with Managers to extend the maturity of the Appellant’s debt from
commercial paper into longer-term debt.
Implementation of the
Guarantee
43. Prior to a corporate
reorganisation in 1988, the Appellant’s issuances of commercial paper and
unsecured debentures had been guaranteed by GE Canada, a Canadian company that
was an indirect subsidiary of [GE] and that carried on industrial businesses in
Canada at the time.
44. [GECUS] began guaranteeing the Appellant’s
issuances of commercial paper and unsecured debentures after the corporate
reorganisation in 1988.
45. Since 1988, including the years under appeal, [GECUS]
unconditionally guaranteed payments due under the Debt Securities issued by the
Appellant.
46. The full text of the [GECUS] guarantee was
printed on the promissory notes issued under the Appellant’s Commercial Paper Program
. . . .
47. Similarly, the full text of [GECUS]’s
guarantee was printed on the notes representing the Unsecured Debentures
. . . .
Credit Ratings
48. In general, a credit rating is a credit
rating agency’s opinion of the general creditworthiness of an obligor, or the
creditworthiness of an obligor with respect to a particular debt security or
other financial obligation, based on relevant risk factors.
49. An “issue rating” is a credit rating agency’s current opinion
of the creditworthiness of an obligor with respect to a specific financial
obligation, a specific class of financial obligations, or a specific financial
program (including ratings on medium term note programs and commercial paper
programs).
50. An “issuer rating” is a credit rating agency’s opinion of an
obligor’s overall capacity to meet its financial obligations and focuses on the
capacity and willingness of an issuer to meet all of its obligations as they
become due.
Credit Rating Agencies
51. S&P [Standard and Poor’s] is a credit rating agency based
in the United States.
52. Moody’s Investors Service (“Moody’s”) is a credit rating agency
based in the United States.
53. The Dominion Bond Rating Service (“DBRS”) is a credit rating
agency based in Canada.
54. The Canadian Bond Rating Service (“CBRS”) was a credit rating
agency based in Canada during the years under appeal. CBRS was acquired by
S&P on or around October 31, 2000.
Credit Ratings in Respect of [GECUS]
55. During the years under appeal, S&P assigned an issuer
rating of AAA to [GECUS], the highest issuer rating assigned by S&P.
56. During the years under appeal, Moody’s assigned an issuer
rating of AAA in respect of [GECUS], the highest rating on Moody’s credit
rating scale.
Credit Ratings in Respect of the Appellant
57. During the years under appeal DBRS assigned to the Appellant’s
Commercial Paper Program a public credit rating of “R-1 high”. According to
the DBRS credit rating scale, commercial paper rated R-1 (high) is of the
highest credit quality, indicating an entity possessing unquestioned ability to
repay current liabilities as they fall due . . . .
58. In 1999, CBRS assigned to the Appellant’s Commercial Paper Program
a public credit rating of “A-1 (High)”. The rating of A-1 (High) was the
highest rating on the CBRS credit rating scale for commercial paper
. . . .
59. S&P and Moody’s assigned credit ratings to the Appellant’s Unsecured
Debentures and Commercial Paper Program during the years under appeal as part
of detailed credit research and ratings reports on [GECUS] . . . .
60. No credit rating agency assigned a public issuer rating to the
Appellant during the years in issue.
The Guarantee Fee
Implementation of the Guarantee Fee
61. Prior to 1995, [GECUS] did not charge the Appellant for the
guarantees of the Appellant’s Debt Securities.
62. In 1995, the Appellant’s Board of Directors resolved to pay a
fee for [GECUS]’s guarantees of the Appellant’s Debt Securities . . . .
63. The Appellant and [GECUS] entered into written agreements
concerning the guarantee fees (“Guarantee Fee Agreements”) . . . .
64. Pursuant to the Guarantee Fee Agreements, [GECUS] agreed to
guarantee the Appellant’s Debt Securities and the Appellant, in turn, agreed to
pay a fee to [GECUS] equal to 1% (or 100 basis points) per annum of the
principal amount of the Debt Securities outstanding from time to time during a
year.
65. Guarantee fees were payable in respect of unsecured debentures
issued on or after April 13, 1995.
66. In respect of commercial paper, guarantee fees were payable in
respect of issuances on or after October 31, 1995 . . . .
Calculation and Payment of the Guarantee
Fees
67. During the years under appeal, the
guarantee fees were calculated on a quarterly basis by the Treasury Department
at [GECUS].
68. The Appellant typically paid the annual guarantee fee to [GECUS]
in the year following accrual.
Tax Treatment of the Guarantee Fees by the Appellant
69. In computing its income for the years in question, the
Appellant deducted the guarantee fees that had accrued in each year, in the
following amounts:
|
Taxation
Year
|
Guarantee
Fees
|
|
1996
|
$30,974,070
|
|
1997
|
$37,149,390
|
|
1998
|
$36,339,640
|
|
1999
|
$17,586,207
|
|
2000
|
$14,378,876
|
70. The Appellant withheld and remitted to the Receiver General for
Canada withholding tax at the rate of 10% from the guarantee fees paid to [GECUS]
in respect of the 1996 to 2000 taxation years, pursuant to paragraph 212(1)(b)
and subsection 214(15) of the Act and Article XI of the Canada‑United
States Income Tax Convention.
The Minister’s Assessments and Reassessments
71. In reassessing the Part I tax payable by the Appellant, the
Minister disallowed the full amount of guarantee fees claimed as deductions by
the Appellant, as follows:
|
Taxation year
|
Guarantee
Fees Disallowed as Deductions in Computing Income
|
|
1996
|
$30,974,070
|
|
1997
|
$37,149,390
|
|
1998
|
$36,339,640
|
|
1999
|
$17,586,207
|
|
2000
|
$14,378,876
|
72. The Minister made the following assessments of the Appellant’s
tax payable under Part XIII of the Act (collectively, the “Part XIII
Assessments”):
|
Notice of Assessment Number
|
Notice of Assessment Date
|
Taxation Year‑End
|
Adjustment to Part XIII Tax
|
|
6240554
|
March 11, 2004
|
December 31, 1996
|
$1,548,704
|
|
6240555
|
March 11, 2004
|
December 31, 1997
|
$1,857,470
|
|
6240556
|
March 11, 2004
|
December 31, 1998
|
$1,816,982
|
|
6240557
|
March 11, 2004
|
December 31, 1999
|
$879,310
|
|
6240558
|
March 11, 2004
|
December 31, 2000
|
$718,944
|
. . .
[3]
The Appellant called 13 witnesses,
seven of whom were qualified as expert witnesses. The Respondent called seven
witnesses, five of whom were qualified as expert witnesses.
SUMMARY OF APPELLANT’S
WITNESSES’ TESTIMONY
Testimony of
Roman Oryschuk
[4]
Mr. Oryschuk
testified that he was the president and CEO of GE Capital Solutions (“GECS”), a
Canadian subsidiary of the Appellant, during the taxation years under review. Previously,
he had run the equipment leasing business for the National Bank of Canada, and he became a GE employee when that business was
sold to the Appellant.
[5]
He was responsible for
the Canadian business operations of the entity. He led a very dynamic sales
team which was able to expand the mid-market equipment leasing business of the
Appellant over the taxation years in question. He was not responsible for
meeting the financing needs of the business. That function was performed by GECUS,
the guarantor of the Appellant’s debt.
[6]
The operations of the
Appellant expanded rapidly over the period in question. The total assets of the
business grew from $1.3 billion as Mr. Oryschuk’s arrival to
approximately $5 billion when he left to take on his new duties in Europe
for a GE affiliate.
[7]
Mr. Oryschuk
testified that, similar to the situation in all of the GE enterprises, he had
operational autonomy provided his operations could meet the financial benchmarks
all GE entities were expected to meet. In the GE culture, an annual review was
performed to determine growth prospects for the business, new markets, etc. If
the Canadian operations failed to meet their financial targets, it was clear in
the witness’s mind that the businesses could be sold or wound down. According
to the witness, this could be done through a sale of the assets backing a
particular loan or leasing portfolio, or of the entire business if the problem
was more general in nature.
[8]
Mr. Oryschuk
further testified the AAA rating obtained by the Appellant had an impact on his
company in that it allowed funds to be obtained at competitive rates.
Testimony of Jeffrey Werner
[9]
Mr. Werner
testified that GE, headquartered in Fairfield Connecticut, was the parent of 12 separate lines
of business, which ran independently of each other. Of these lines of business,
11 were considered part of the industrial businesses, e.g. aircraft engines,
major appliances, plastics, etc. The remaining business line was the financial
services business.
[10]
Within the financial
services business, GE held all the shares of GECSUS, a holding company for the
entities in this particular business line. GECSUS, in turn, held shares in
GECUS, GE Global Insurance Holding Corp. and Employer’s Reinsurance Corp.
[11]
The main purpose of
GECUS was to fund the operations of affiliated companies by issuing commercial
paper and unsecured debt instruments. After raising the funds needed, GECUS would
provide those funds to the affiliates through intercompany transfers. To avoid
liquidity problems, GECUS maintained backup lines of credit in an amount equal
to 50% of its outstanding commercial paper. GECUS handled all the treasury
functions for all of the businesses within the GE Capital group, including the
Appellant.
[12]
After the
reorganization in 1988, the Appellant became an affiliate of GECUS. The
guarantees that were provided by the Appellant’s former Canadian parent (GE
Canadian Holdings Ltd.) remained with the Canadian parent and GECUS began
to provide guarantees to secure the Appellant’s debt around April 1989. There were no
guarantee fees charged by GECUS from 1988 until 1996, the first taxation year
under appeal.
[13]
Although no fee was
charged initially, as part of a transfer pricing review Mr. Werner was
asked to make a recommendation to Mr. Jim Parke, CFO of GECUS, as to what
guarantee fee the Appellant should be charged. Mr. Werner ultimately
recommended a guarantee fee of 100 basis points (1%) of all new debt issued.
[14]
In determining the
guarantee fee to be charged, Mr. Werner testified, he understood the fee had
to be an arm’s length price.
Mr. Werner compared the rates that the Appellant would be able to obtain
if it could borrow with the guarantee (investment grade rating) with the rate at
which it could borrow without the guarantee (a non-investment grade rating). This method is
referred to as the yield curve approach and it reflects the costs of borrowing
money given various maturities and different credit ratings. The differential
or spread was determined to be between 100 and 300 basis points (1% to 3%).
[15]
The guarantee fee of
100 basis points (1%) was ultimately recommended by Mr. Werner since he
believed it accurately represented the benefit the Appellant had been enjoying.
Moreover, had there been arm’s length negotiations, the Appellant would have kept
some of the benefit for itself.
[16]
Mr. Werner
testified that he did not believe the Appellant would have been rated AAA
without the guarantee.
He denied GECUS would necessarily have supported the Appellant’s operations; more
precisely, he claimed that GECUS could have just walked away if there was no
explicit guarantee in place.
[17]
During
cross-examination, Mr. Werner conceded that its AAA rating was very
important to GECUS.
GECUS, as an unregulated financial services company, did not have access to low‑cost
borrowings made available through deposits, to which banks do have access.
Accordingly, GECUS and the Appellant were principally reliant on credit
markets.
In the commercial paper market, they had to constantly roll over commercial
paper.
[18]
Mr. Werner
admitted that GE valued its reputation, as stated in its public filings. He
also admitted the Appellant used some of the same dealers and underwriters as
those that are used by GE subsidiaries throughout the world to issue its
commercial paper and bonds.
Testimony of Laurence Booth, Ph.D.
[19]
Dr. Booth has a
bachelor’s degree in science and economics from the London School of Economics,
an M.A., an M.B.A. and a Ph.D. in business administration from Indiana University.
He is a professor of finance at the University of Toronto,
holds the CIT Chair at the Rotman School of Management, and is in charge of the
investment banking track in the MBA program at the University of Toronto.
His experience includes acting as an expert witness in financial and capital
market matters, mainly with regard to regulated industries. Specifically, in
the case at bar, Dr. Booth was asked to give an opinion on:
(a) whether it was prudent
and reasonable for the Appellant to obtain a guarantee from GECUS; and
(b) what the market value
of that guarantee was.
[20]
Dr. Booth
acknowledged that he is not an expert in establishing credit ratings and stated
that he was not asked to do so. In fact, he noted that most U.S. or multinational parents guarantee the debt
obligations of their Canadian subsidiaries.
[21]
Unregulated financial
services companies usually operate without a captive retail deposit base. Conversely,
regulated banking institutions have extensive retail networks and generate a
significant amount of their funds from customers’ deposits which, in turn, are
covered by the Canada Deposit Insurance Corporation. This allows banks to have
a large source of low‑cost and stable funds. Without a guarantee,
financial services companies such as the Appellant would be competing at a
significant disadvantage. According to Dr. Booth, the Appellant would not,
in the absence of the guarantee, have been able to raise capital as it did.
[22]
Dr. Booth maintained
that the guarantee also offered other advantages to the Appellant; for
instance, it did not have to pay placement fees when issuing commercial paper
through the dealer network. In addition, the Appellant avoided the need for
backup lines of credit, which would have been required if the debt was not
guaranteed. Typically, 100% of an issuer’s commercial paper must be supported
by backup lines of credit to guard against the refinancing risk that arises
during periods of financial instability where investors look to government-issued
treasury bills and the like. The Appellant did not have to negotiate backup
lines of credit because GECUS was viewed as the ultimate creditor and had
arranged for the availability of such credit facilities. That arrangement
coupled with the explicit guarantee allowed the Appellant to piggyback on the
sterling credit rating of GECUS.
[23]
Dr. Booth
testified that he was of the opinion that the Appellant could not have received
a AAA rating or an R-1 high commercial paper rating without an explicit
guarantee from GECUS.
The guarantee and the rating allowed the Appellant’s commercial paper to be
virtually risk free, similar to treasury bills. The guarantee
meant the Appellant could raise significant sums of money several times a day,
over the telephone, from a variety of investors. Dr. Booth
admitted that the GE name alone would “notch up” the credit rating; however, he
believed it extremely unlikely that this would have got the Appellant a credit
rating sufficient to permit access to the commercial paper and swap markets in
the volumes the Appellant needed to finance its Canadian operations.
[24]
Dr. Booth
acknowledged the capital markets would be shocked if GECUS allowed its Canadian
subsidiary to default, assuming it had not guaranteed the Appellant’s indebtedness.
However, Dr. Booth advanced two circumstances in which the parent could allow
a subsidiary to default: a single episodic shock that causes a huge loss of
money or a gradual deterioration of the business. Dr. Booth
pointed out that it was the economic value of the subsidiary that was
important, not whether it had the same name as, or was considered a “core” or
“strategically important” subsidiary of, GECUS. Dr. Booth asserted that if
the parent’s economic incentive is to walk away, then the parent will walk away,
thereby changing any classification, e.g. core or strategically important, that
the subsidiary may have.
Testimony of Brian Neysmith
[25]
Mr. Neysmith, now
retired, was one of the co‑founders of CBRS in 1973. After the company
was sold to S&P in 2000, he remained with S&P until January 2003. His main
responsibility, after CBRS was sold, was to “merge the two companies together
to harmonize the rating criteria [and] the rating methodology”. Mr. Neysmith
was the person who “signed off” on, and authorized the issuance of, new
ratings.
[26]
During the relevant
period, CBRS’s share of long‑term debt and commercial paper rating in the
Canadian medium-term market was significant; for example, in the commercial
paper market CBRS held between 50% and 60% of the rating business.
[27]
Mr. Neysmith stated
that both CBRS and DBRS approached credit analysis from the same fundamental
basis: looking at the ongoing financial condition of the company and its
history.
Even though the ratings were similar in about 75% of the cases, the companies
and issuers would usually request both: CBRS and DBRS.
[28]
Mr. Neysmith
testified that CBRS would not have given the Appellant a AAA rating without
GECUS’ guarantee. According to the witness, CBRS would have rated the Appellant
lower than A+ or A-1 high.
Testimony of Mark Fidelman
[29]
Mr. Fidelman is an
economist currently employed as a tax director at Deloitte Tax LLP in the U.S. His expertise pertains to transfer pricing and
insurance industry pricing.
[30]
Mr. Fidelman used
a model designed for pricing insurance products for the purpose of determining
an arm’s length price for the guarantee.
[31]
According to the
witness, a debt guarantee can be viewed as a form of credit insurance.
Financial guarantees are sold as insurance products to improve creditworthiness
with respect to both public-sector and private‑sector debt. Examples of
public entities in the U.S. availing themselves of financial guarantee
insurance on debt issuances are municipalities with regard to economic
development and general revenue bonds.
[32]
According to
Mr. Fidelman, the stand-alone credit rating of the Appellant during the
valuation period was in the range of BB to BB+ under an insurance pricing
model.
[33]
According to the
witness, an insurer would not take into account the benefit of an “implicit
guarantee” from a subsidiary’s parent when pricing insurance. As an insurer
would expect to be asked to pay out under credit default insurance if the
debtor defaulted, the insurer would not expect the debtor’s parent to pay out
on the basis of implicit support and, as a result, would take no account of an
implicit guarantee when establishing an appropriate premium for the product.
[34]
Asked to comment on a
Respondent’s expert, Dr. Saunders, Mr. Fidelman asserted that the expected
loss cost is just one component of the total guarantee fee charged that would meet
the requirements of the arm’s length principle, and that the other major
component is a return on risk capital.
Mr. Fidelman disagreed with the calculation by Dr. Saunders that was based
on the Basel II regulatory regime. In the first place, Basel II was
not in place during the years under study; rather, Basel I
was applicable during that period.
[35]
Basing his opinion on
the insurance pricing methodology, Mr. Fidelman concluded that the 1% fee
charged the Appellant by GECUS does not generate the return on capital an
insurer would demand for an insurance-based guarantee. On this basis, the fee
paid by the Appellant does not exceed what would be charged in an arm’s length relationship.
Testimony of John Frederick Coombs
[36]
Mr. Coombs is a
banker holding the office of vice‑chairman of TD Securities and senior
vice-president of TD Bank Financial Group, head of Europe and Asia Pacific. This witness
was qualified as an expert in banking and credit matters. He was not
qualified as a transfer pricing or credit rating expert.
[37]
Mr. Coombs
concluded the Appellant would not have been able to borrow the amount of funds
it did in the Canadian commercial paper market without GECUS’ guarantee.
[38]
Mr. Coombs
testified the Appellant’s debt-to-equity ratio was higher than that of other
independent companies in the market. The Appellant’s ratio was between
10 to 1 and 12 to 1 whereas other unregulated financial
institutions had ratios between 5 to 1 and 8 to 1. On the
basis of this factor alone, Mr. Coombs believed the Appellant would not have
been rated investment grade.
[39]
Without a formal
guarantee from the parent company, Mr. Coombs asserted, Canadian banks
would have extended only limited credit facilities to the Appellant. Those
facilities would have been short-dated (in the 3- to 12-month range) foreign
exchange, derivative and operating credit facilities in amounts ranging up to
perhaps $100 million in total.
The inference made by Mr. Coombs was that the Appellant would not have
been able to negotiate large enough backup lines of credit to support its
commercial paper program without the guarantee by its parent and, as a result,
the Appellant would not have been rated investment grade.
[40]
Mr. Coombs
maintained that his opinion would not be influenced by the fact that a credit rating
agency may have been willing to give the issuer an investment grade rating.
Banks generally use the public debt ratings as a guideline or ceiling, but
assign internal ratings based on their own fundamental analysis. TD Securities
frequently assigns internal risk ratings below those determined by rating
agencies. Banks would provide some accommodation, but Mr. Coombs did not
believe the kind of accommodation given would have been sufficient for the
Appellant to operate its business during the relevant period.
[41]
In his opinion, without
the guarantee, the Appellant would not have been able to obtain from banks the
backup lines of credit required in order to get a DBRS or, at the time, CBRS,
investment grade rating.
[42]
Mr. Coombs
dismissed the implicit guarantee argument, reasoning that if the parent would
never let the Appellant fail, Canadian banks would expect the commitment to be
backed up by a formal guarantee.
Testimony of William John Chambers, Ph.D.
[43]
Dr. Chambers has
been a professor at Boston University since 2005. He received a B.A. from College of Wooster in 1968 and an M.A., an M.Phil. and a Ph.D. in
economics at Columbia University, the
latter degree having been conferred in 1975.
He began his career with S&P in 1983 and held a number of senior positions
with the company until he left in 2005.
[44]
Dr. Chambers was
asked to opine on the credit rating of the Appellant for the taxation years
under review, assuming its indebtedness was not guaranteed by GECUS. He was
instructed to use for this purpose the S&P rating criteria and methodology
applicable in the years in question.
[45]
Dr. Chambers explained
that in preparing a credit rating for a company that is a subsidiary of
another, S&P first determines the stand-alone creditworthiness of the parent
and the subsidiary. At this stage, Dr. Chambers asserted, the rating
agencies take into account the relationship between the parent and the
subsidiary in terms of the management services and expertise provided by the
former, and their common name.
However, neither S&P nor Moody’s would take into account the possibility
that the parent might inject funds into the subsidiary or provide financial
support through a different means.
[46]
Dr. Chambers
concluded that the stand-alone creditworthiness of the Appellant, using the
S&P rating criteria, would have been a single B+ or a BB- during the
relevant period.
[47]
There was much
confusion during the trial regarding the distinction between a stand-alone and
status quo rating. My understanding was that both of these ratings look at the
subsidiary as a separate business as opposed to part of an integrated group.
However, the distinction between the two is that the status quo rating takes
into account the benefits of the common name, the parent’s management team and
existing business arrangements, such as intercompany loans. The stand-alone
rating abstracts from these conditions. Because Dr. Chambers takes these
elements into account, his first step results in a status quo rather than stand-alone
rating.
First Step: Analysis of the Appellant on a Stand-Alone/Status
Quo Basis
[48]
Dr. Chambers assigned
a B+ to BB- rating to the Appellant for various reasons. The Appellant was a
profitable entity, growing rapidly in a very stable marketplace. Yet, rapid
growth can be a red flag for financial institutions. The Appellant was thinly
capitalized and its degree of leverage was quite high. Further, its
profitability was also decreasing during the period. Although the Appellant had
reduced its leverage and was growing rapidly, it did not seem able to
continually generate profits or increasing profits. Dr. Chambers noted the
Appellant’s return on equity rose during 1995 and 1996, and then started to
deteriorate.
[49]
Dr. Chambers
looked at the competitive environment which was and continues to be very
intense. The market share of large banking institutions declined, but overall the
sector was growing rapidly: about 10% per year during the relevant period.
[50]
In that competitive market,
the Appellant was a new entrant. The GE group had a clear mandate to grow, which
translated into numerous acquisitions.
Overall, their growth was about 20% per year compounded.
[51]
Dr. Chambers
believed the Appellant’s rating would have been stronger if the Appellant had
maintained a debt-to-equity ratio of 7 to 1 or 8 to 1,
similar to GECUS’.
[52]
This element was weighted
heavily by Dr. Chambers in establishing the stand-alone/status quo
rating. While the company was profitable, its fast‑paced growth could
lead to management problems. Further, the increase in loan loss provisions and
the degree of leverage made the company susceptible to additional risks.
Second Step: Factoring the Parent-Subsidiary
Relationship into the Stand‑Alone/Status Quo Rating
[53]
Dr. Chambers
testified that once all of the relevant factors are properly weighed by the
rating analyst, the analyst will seek to rank the subsidiary in a spectrum that
ranges from entities considered “core” on one end to “independent” on the
other. “Core”, in rating agency nomenclature, generally describes a subsidiary
that would be supported by the parent in virtually all circumstances. “Core”
tends to connote that the subsidiary constitutes a critical part of the group’s
business, represents a large proportion of existing business, and enjoys a
large market share in terms of the central products or markets of the group. A
subsidiary is more likely to be deemed “core” if its financial performance and
growth exceed that of the aggregate business; thus, support is very unlikely to
be required in this circumstance. In other words, the subsidiary is a key part
of the organization; essentially, the organization could not function in its
present form without that entity being present. Not
surprisingly, subsidiaries considered “independent” are not expected to benefit
from parental support as they can be sold or closed down without any impact on
the financial well‑being of the group as a whole.
[54]
In situations where the
parent is deemed to be financially stronger and more creditworthy than its
subsidiary, Dr. Chambers acknowledged, two approaches are applied to
determine the subsidiary’s final credit ratings. These methodologies are
referred to as the top-down and the bottom-up approaches.
[55]
Under the top-down
approach, the starting point is the parent’s superior credit rating. Dr. Chambers
believed one of the problems with the top-down approach is that the parent and
subsidiary come to the marketplace when they both are generally doing well. However,
ratings are also about looking at what can happen during periods of financial
stress. According to the witness, the bottom-up approach is more sound because
it starts from the stand-alone or status quo rating and is adjusted to take
into account the amount of support that can reasonably be anticipated in the
circumstances.
[56]
Dr. Chambers asserted
that a weak entity owned by a strong parent usually, although not always, will
enjoy a stronger rating than it would on a stand-alone basis. Assuming the
parent has the ability to support its subsidiary during a period of financial
stress, the spectrum of possibilities still ranges from ratings equalization at
one extreme to very little or no help from the parent’s credit strength at the
other. Where there is a large gap, rating agencies will demand more evidence of
likely support.
[57]
In his report,
Dr. Chambers canvassed the factors used by S&P to bridge the rating
gap, which include strategic importance, percentage ownership, management
control, shared name, domicile in the same country, common source of capital,
financial capacity for providing support, significance of amount of investment,
investment relative to amount of debt, the nature of other owners, management’s
stated posture, the track record of the parent company in similar circumstances
and the nature of potential risks.
[58]
According to the
witness, factors are not equally weighted in the evaluation. In general,
economic incentive is the most important factor on which to base judgments
about the degree of linkage that exists between a parent and a subsidiary.
[59]
Dr. Chambers
looked at the factors one by one in light of the GE organization as a whole. For
example, operating in Canada was strategically important for GECUS, as
it was part of their overall plan for international expansion. At the same
time, Dr. Chambers noted that the Appellant’s assets represented a very
small amount of GECUS’ consolidated assets. For that reason, Dr. Chambers
could not infer that the Appellant had a lot of strategic importance in the overall
scheme of things in the GE organization.
[60]
Dr. Chambers
recognized the Appellant was a wholly-owned subsidiary and that a
parent will be more inclined to support a wholly-owned subsidiary than one in
which it has a minority’s stake.
[61]
The witness noted that
while there was overall management control from the top-down, on a day-to-day
basis there was a lot of operating discretion left to the local operation. The Appellant
did not perform any essential or key services for its parent company.
[62]
The witness observed that
the Appellant and GECUS did not have the same source of capital. Both companies
were tapping the wholesale capital markets for their funding, but the specific
instruments used were different. The Appellant focused primarily on the
Canadian market and the European markets, whereas GECUS and the other subsidiaries
borrow on the U.S. and international markets.
[63]
Midway through the
spectrum, the “strategically important” category runs the gamut between the two
extremes (“core” and “independent”). Some aspects of the business, i.e. a
particular product or geographical location, are important as a growth engine
for the organization or to stabilize the group.
[64]
Dr. Chambers
concluded this part of his analysis by determining that the Appellant would be
ranked as an “independent” subsidiary of GECUS by the rating agencies, meaning GECUS
would be expected to provide little financial support to the Appellant in times
of financial stress, if one assumes, that is, that its debt obligations were
not guaranteed by GECUS.
Credit Rating of the Appellant Without the Guarantee
[65]
Concerning the final
rating of the Appellant, Dr. Chambers stated that, without an explicit
guarantee, the stand-alone rating might be raised one or two notches on the
basis of its being within the GE Capital group of companies. Thus, if the
Appellant’s stand-alone rating was determined to be B+, its final rating would
reasonably have been BB- or BB. If the stand-alone rating was determined to be
BB-, the final assigned rating to the non-guaranteed obligations would be BB or
BB+.
[66]
Dr. Chambers’
opinion was influenced by the fact that all of the Appellant’s indebtedness was
guaranteed from 1988 to 1995. Assuming all new debt issues, beginning in 1996,
would not have been guaranteed, Dr. Chambers believes the rating agencies
would have demanded a clear articulation of the support that could be expected
from the parent.
In practice, the rating agency would probably get the company on the phone and ask
for a clear statement concerning the issue.
In Dr. Chambers’ opinion, it would have been difficult for GECUS to
convince a ratings committee that nothing had changed after removing the
explicit guarantee.
Testimony of John Campbell Hull, Ph.D.
[67]
Dr. Hull is the
Maple Financial Group Professor of Derivatives and Risk Management at the
Rotman School of Management at the University of Toronto. He holds a B.A.
and an M.A. in mathematics from Cambridge University in England,
an M.A. in operational research from Lancaster University in England and a Ph.D. in finance from Cranfield University in England.
[68]
Dr. Hull was asked
to provide an opinion on the value of the guarantee in issue by analogy to the
price of a credit default swap (“CDS”). He valued the guarantee by using the yield
approach, which consisted of analyzing the spread between AAA- rated bonds and
bonds that are an average of single B and BB, the credit rating Dr. Hull
was asked to assume for the Appellant in the absence of an explicit guarantee.
[69]
In Dr. Hull’s
opinion, a CDS is a form of financial guarantee. In a CDS, there are two
parties to the transaction, the provider of the protection, which often is an
insurance company or financial institution, and the buyer. Typically the buyer
of protection holds debt securities in a reference entity and seeks protection
from credit defaults. In the event of a default by the reference entity, the
provider of protection makes the buyer whole by paying up to the face value of
the bond.
[70]
For purposes of his
analysis, Dr. Hull assumed the guarantee would be in place for as long as
the Appellant stayed in business in Canada, because if
the guarantee was removed, the Appellant would be unable to issue new
commercial paper to obtain the funds needed to pay the maturing commercial
paper.
[71]
Nonetheless, in his
calculation of the spread between the credit rating of the Appellant with a
guarantee and its rating without a guarantee, Dr. Hull uses the yields on
bonds with maturities of five and 10 years.
[72]
The witness concluded
that the overall spread was about 352 basis points between a AAA rating and
the B+ to BB- rating assumed for the Appellant in the absence of an explicit
guarantee. As a result, the value of the explicit guarantee is approximately
1.83% based on a BB+ to BBB- credit rating range.
Testimony of Stephen Cole
[73]
Mr. Cole is an
accountant and senior partner at the firm Cole & Partners. He is a fellow
of the Institute of Chartered Accountants of Ontario, a fellow of the Canadian Institute
of Chartered Business Valuators and a member of the ADR Institute of Canada Inc.
[74]
Mr. Cole was asked
to explain the conceptual issue of valuation in the context of determining an
arm’s length price for the guarantee. He commented on whether the relationship
between the Appellant and GECUS would influence the price at which a third-party
guarantor would be willing to get involved.
[75]
According to
Mr. Cole, the guarantor would start with the concept of arm’s length
price. The arm’s length price is the amount determined in accordance with the
arm’s length principle that would have been reasonable in the circumstances if
the parties to the transaction had been dealing at arm’s length.
[76]
Mr. Cole described
a notional negotiation in his report. The market dynamic, including a buyer and
a seller, is integral to the definitions of fair market value and arm’s length
price. Mr. Cole portrayed the buyer’s and the debtor’s perspectives and
then the perspectives of the arm’s length providers of the guarantee. The essence of any
negotiation is that the parties can find a middle ground that is mutually
satisfactory. The objective of the buyer, here the Appellant, is to find
insurance or a guarantee at the lowest possible price. Conversely, the provider
would like the highest possible price.
The market needs to be composed of many providers and there were numerous parties
around the world with similar characteristics, such as deep‑pockets,
substantial size and sophistication in the financial sphere, who could all have
provided the guarantee.
[77]
Mr. Cole reached
the conclusion that the range of arm’s length and fair market value prices for
the guarantee is clearly above 1%.
Mr. Cole arrived at a 0.85% fee only with respect to a situation where the
guarantor perceived the Appellant as being a AAA credit risk and there was no
explicit guarantee. Yet, Mr. Cole does not believe that it was realistically
possible for the Appellant to have obtained a AAA rating on its own.
[78]
Mr. Cole noted
there was no benefit to GECUS in taking a dividend from the Canadian
subsidiary; had they taken a dividend, they would simply have been increasing
debt in the Canadian company and paying down debt in the U.S. company, but its aggregate access to capital would have
been identical on a consolidated basis.
There would have been no benefit to the parent either if the Appellant had been
capitalized differently because GECUS’ debt‑to‑equity ratio is determined
on a consolidated basis and because its access to capital is also determined on
a consolidated basis.
In summary, GECUS’ borrowing capacity was linked, inter alia, to its
consolidated debt‑to‑equity ratio.
[79]
Mr. Cole justified
the reasonableness of both his guarantee fee and the guarantee fee charged by
GECUS by looking at the proportion of the guarantee fee to the profits earned
by the Appellant in each of the taxation years at issue. According to the
witness, a 1% guarantee fee was in the range of 33 to 40% of the aggregate
pre-tax and pre-guarantee fee profits of the Appellant. Mr. Cole believed
GECUS would have been entitled to a return of between 15% and 20% if it had
invested roughly $275 million of additional capital in the Appellant to
improve the Appellant’s creditworthiness. This percentage equates to roughly
$69 million of profit, which is very close to the average guarantee fee
that was in fact received.
[80]
In cross-examination,
Mr. Cole confirmed his opinion that the Appellant could not have obtained a
AAA rating without the guarantee. The witness believed an explicit guarantee
was necessary in order to enable the Appellant to execute its business plan, which
was based on achieving the lowest cost of capital.
Testimony of Rowland Alexander Lewis
[81]
Mr. Lewis was
called upon by the Appellant to provide evidence on how the treasury
departments of large corporate multinationals would view the concept of
implicit support when pricing Canadian‑denominated commercial paper used
to invest short-term Canadian‑denominated funds. From 1980 to 2000,
Mr. Lewis worked in the treasury department of Texaco Canada and Heddington Insurance, a captive insurer of
Texaco.
[82]
Mr. Lewis
testified that in his role of assistant treasurer he would invest in commercial
paper based on investment guidelines, which included safety of principal,
liquidity and foreseeable profit.
Mr. Lewis defined implicit support as an assumption by third parties that,
in the event of an issuer’s default, the parent company of the issuer would
step in and keep everybody whole.
Texaco would not have assigned any value to implicit support during the
relevant period because all its investments in short-term paper had to be made in
such a way as to ensure safety of principal.
Given a choice between an unguaranteed debt issuance of the Appellant and a
debt issuance of GECUS or GE, Texaco would buy the latter.
Testimony of David Victor Daubaras
[83]
Mr. Daubaras has
held the position of vice-president, tax, with the Appellant since 1995.
[84]
Mr. Daubaras did not
believe the Appellant was simply an extension of GECUS. According to the
witness, the U.S. would provide the Appellant with a
financial target and it would be up to the Canadian management of the Appellant
to achieve that target.
Testimony of Bruce Bennett
[85]
Mr. Bennett was
the associate general counsel in the treasury division of GECUS between 1992
and 1998.
The witness was responsible for overseeing a small group of lawyers who
supervised and implemented GECUS’ debt issuance, insurance and derivative
transactions.
[86]
Mr. Bennett
testified that the commercial objectives of GECUS’ treasury department were to
fund GECUS’ consolidated business activities at the lowest possible cost while
at the same time adhering to the risk management guidelines established by
senior management.
These guidelines required that liabilities match the income stream of the
assets they funded. For example, an asset that generated a short-term income
stream would be supported by short‑term borrowing.
[87]
Mr. Bennett
acknowledged that GECUS valued its AAA rating because none of its competitors
were able to borrow at the same low rate. The witness’s understanding was that
GECUS’ guarantee allowed the Appellant to borrow at the lowest interest rate.
From a tax standpoint, the guarantee enabled the Appellant to borrow without having
to bear the cost of the Canadian withholding tax. If GECUS borrowed the funds
directly and then loaned them to the Appellant, withholding tax would have been
due on the interest payments. The guarantee also provided ease of execution. From
the investor’s standpoint it was as if GECUS was borrowing the funds directly.
Fungibility, from a credit risk standpoint, was achieved between the commercial
paper issued by GECUS and that issued by the Appellant because of the explicit
guarantee. In summary, the explicit guarantee meant that the Appellant’s debt
holders benefited from GECUS’ lower credit risk much in the same way that they
would have if the funds had been loaned to GECUS in the first instance and then
loaned by GECUS to the Appellant.
SUMMARY OF RESPONDENT’S WITNESSES’ TESTIMONY
Testimony of Stephen Allan Mitchell
[88]
Mr. Mitchell is
currently employed as an investment banker by RBC Capital Markets, a
wholly-owned subsidiary of the Royal Bank of Canada
(“RBC”). In 1995, he was senior manager of corporate banking at RBC.
[89]
From 1995 to 2000, RBC
had business dealings with the Appellant. The nature of these dealings
encompassed cash management services and credit facilities. His current
employer, RBC Capital Markets, has an ongoing relationship with the Appellant.
[90]
Mr. Mitchell is the signatory
of a November 7, 1995 letter apparently written in response to an October 10,
1995 letter, both marked as Exhibit R‑9A. The October 10
letter sought information regarding the pricing of loans to be made to Canadian
subsidiaries of GE and GECUS. Specifically, this letter involved an inquiry by
the Appellant concerning GECUS’ borrowing capacity on a stand-alone basis and concerning
the pricing of a $2 billion credit facility without an express guarantee
from the parent.
[91]
Under
cross-examination, Mr. Mitchell testified that, as written in the third
paragraph of the November 7, 1995 letter, considering the structure of the
Canadian balance sheet at that time and the inherent leverage and debt coverage
ratios, the Canadian company would come in below the BB level. However, the
caveat to this is that the letter dated November 7, 1995 did not involve any
transaction between RBC and the Appellant; rather, it provided market feedback
only and, accordingly, one would not have sought credit approval or credit
adjudication within RBC for the purpose of issuing this type of letter.
[92]
In the same letter,
various GE Canadian subsidiaries were identified as requiring the parent
company’s guarantee before a commitment of bank credit facilities could be
envisaged. Mr. Mitchell testified to the effect that the fact that they
were GE subsidiaries was insufficient in and of itself to warrant a credit facility
in the absence of a formal guarantee.
He went on to testify that the suggestion in the said letter was that at the B rating
level the fully drawn cost would be 250 basis points, plus LIBOR. Mr. Mitchell
also agreed that the Appellant could not be considered a AAA-rated company.
Testimony of Kevin Charles Clark
[93]
Mr. Clark was, in
1995, a credit officer in 1995 at Scotiabank who was responsible for covering
the Connecticut territory. His objectives involved building
relationships with the bank’s borrower clients. GECUS and GE were part of his
portfolio and were clients of Scotiabank.
[94]
Mr. Clark was unclear
as to the purpose underlying the inquiry from GECUS; generally, as in the letter
of July 24, 1995 for example, he was asked about the cost of borrowing
funds in Canada both with and without support from the
parent company.
He believed at the time that the letter would lead to a lending relationship
and also that the request for credit was for general corporate purposes.
[95]
Corporations like GECUS
or the Appellant, whose businesses are large and diversified, frequently borrow
for what Mr. Clark called general corporate purposes, which
provides them with flexibility in their use of the funds.
[96]
The Bank of Nova Scotia
would have been prepared, in 1995, to take on the responsibility of putting
together a $2 billion unguaranteed commitment to the Appellant. This
would have been achieved through syndication, that is, the bank would have been
prepared to try to arrange such financing for the Appellant.
[97]
In the circles of
syndicated lending, having the responsibility of arranging a facility for a
well‑respected name in the marketplace would have been viewed as an
attractive opportunity, even though the facility was not for GECUS
specifically. It would have given the Bank of Nova Scotia so‑called “bragging
rights” in the syndicated debt league circuit.
[98]
In the evaluation of
the request for an unguaranteed loan, the attractiveness of the GE name and of the
affiliation with GE was relevant to Mr. Clark’s analysis, in light of the
fact that GECUS was among the largest users of commercial paper in the U.S.
[99]
In paragraph 3 of his
letter (Exhibit R-9B), Mr. Clark referred to “ownership”. According to
the witness, the term “ownership” referred to borrowing with a guarantee. He
would give different rates depending on whether there was or was not ownership.
Thus, the Appellant would have been able to borrow at a preferential rate if
there was a guarantee from GECUS.
Testimony of Anthony Saunders, Ph.D.
[100]
Dr. Saunders
received a B.S. in economics, an M.S. in economics and a Ph.D. in economics,
all from the London School of Economics.
Dr. Saunders, a professor of finance at the Stern School of Business, New
York University, was qualified as an economist specializing in credit risk
measurement and in analysis and valuations relating to debt guarantees based on
creditworthiness.
[101]
In his report, Dr. Saunders
did not address the credit rating of the Appellant as a stand-alone entity; he
considered that it was obvious the Appellant was a core or strategically
important company in relation to its parent, GECUS.
[102]
As a preliminary
procedural matter dealing with the admissibility of Dr. Saunders’ rebuttal
report, the Appellant asserted that this report was not rebuttal per se, but served
rather to fill in gaps in Dr. Saunders’ initial report. I mentioned that
my decision on this would be reserved until my final judgment. The Appellant did
not succeed in convincing me that, on a balance of probabilities, the rebuttal
aspects of Dr. Saunders’ rebuttal report were not within the confines of
rebuttal evidence. As I see it, Dr. Saunders addressed the stand‑alone
approach in his rebuttal report in order to counter the approach adopted by the
Appellant, an approach with which he did not agree in the circumstances. While
it stands as true that he recognized that a stand‑alone analysis is part
of the methodology employed by Moody’s and S&P, it is also
clear that Dr. Saunders did his analysis without employing that methodology
because he considered this step redundant. In short, Dr. Saunders believed
it was unnecessary to perform a stand-alone analysis because the conclusion was
obvious in the circumstances. Therefore, I do not agree with the Appellant’s
contention that such an analysis was in fact part and parcel of the methodology
that Dr. Saunders purported to apply. In this light, I view the rebuttal report
authored by Dr. Saunders as legitimate and viable rebuttal per se.
[103]
Dr. Saunders used
three steps to evaluate the reasonableness of the 1% fee. First, he identified
how the rating agencies would rate the Appellant’s debt. Second, having established
that rating, he calculated an appropriate fee that would be charged by a third‑party
guarantor. Third, he looked at the issue of whether the 1% fee was excessive in
terms of the risk-adjusted return on capital (“RAROC”).
[104]
Dr. Saunders
identified three approaches to creditworthiness measurement: (i) expert
systems, (ii) traditional rating systems, and (iii) quantitative model approaches.
[105]
The expert system is
purely qualitative and is based on characteristics of the borrower, namely:
character, capital and capacity to repay, inter alia. The problem with
expert systems is that they are purely subjective. Where the expert systems differ
from quantitative systems is not so much in identifying the factors, but in
weighing the factors in a consistent fashion.
[106]
As for traditional
credit rating models, they use both qualitative expert systems and quantitative
systems. These models focus on longer‑term default risk.
[107]
The third model, the
quantitative model, is generally based on financial market data, such as stock
prices, bond prices and CDS spread, inter alia. Dr. Saunders
testified that, during the period under assessment, investors, whether
institutional or retail, relied most heavily on the traditional rating models.
This method was employed by Dr. Saunders in his report.
[108]
In summary, Dr. Saunders
concluded that the Appellant was a core subsidiary of GECUS at the relevant time
and would have been rated AAA. Alternatively, he concluded the Appellant could
be classified as having been strategically important, rather than of core
importance, to GECUS, in which case it would have been rated AA. He determined what
the fee would have been for a third-party guarantee given a AA rating, as
opposed to a AAA rating. On this basis, his conclusion was that the fee would have
been between 15 and 24 basis points over the relevant period. When looking at
the risk-adjusted return on capital, Dr. Saunders concluded the 1% fee
produced a very high risk-adjusted return on capital, much higher than what is
called the “hurdle rate”. In RAROC modeling, the “hurdle rate” is the return
on equity required by stockholders before committing to an investment.
[109]
Dr. Saunders contended
that Dr. Chambers did not notch up enough in taking into account implicit support
by the parent company. He further stated that the importance of reputation is greater
for unregulated financial institutions than for regulated ones, as the latter
have the luxury of relying on regulators to come in to clean things up and
constrain their lending activities. In his view, unregulated financial
institutions only have their reputation to enable them to preserve and enhance the
market’s confidence in them, and the notching up should reflect this.
[110]
Dr. Saunders
believed there would have been an enormous cost to GECUS if it had let the
Appellant fail:
I mean, as an economist, as I said, at the end of the day, it is a
simple cost‑benefit question. The cost is enormous. You have a company,
GEC Capital Corporation [GECUS], issuing $2-trillion or more commercial paper
and billions of dollars of bonds. If it could even possibly lead to a one- or
two‑notch downgrade, that would be an enormous increase in its cost of
debt.
[111]
Five factors led Dr. Saunders
to conclude that the rating agencies would likely have viewed the Appellant as
a core subsidiary of GECUS: (i) the specialness, significance and value
attaching to the AAA rating; (ii) the branding aspect of having the same name: investors
are more likely to run from a parent that has the same name as a subsidiary
that fails; (iii) the high degree of financial and managerial integration; (iv)
the longevity of the subsidiary: the Appellant has been in Canada close to 40 years; and (v) the size of
the subsidiary (the Appellant) in relation to the Canadian capital market.
[112]
Dr. Saunders did
not place much weight on the fact that the Appellant is domiciled in Canada and is independent of GECUS. He remarked that with
NAFTA, in a free-trading association between Canada and the U.S., there are strong economic links that reduce the
significance of the separate domiciles of the Appellant and GECUS. Moreover, he
also thought that the Appellant, with a 12 to 1 debt-to-equity ratio,
was well‑capitalized relative to chartered banks, for which an 18 to 1 debt-to-equity
ratio is considered adequate —
bearing in mind that we are dealing with core tier-one equity.
[113]
To calculate the
difference between what a AA subsidiary and a AAA company might pay, Dr. Saunders
reviewed the yield spread between a AAA and a AA company, relying on Canadian
data. There were no real AAA Canadian corporate bonds for him to use, just
Canadian government bonds, which means that the spread would actually be bigger
given that AAA corporate bonds tend to yield slightly higher than AAA
government bonds.
While the spread has a built-in premium for liquidity, because it does not measure
default risk only, Dr. Saunders believes he proceeded conservatively as he
assumed the liquidity spread to be zero, thereby leaving the whole spread for
default risk.
[114]
Dr. Saunders also
performed a mortality analysis; his conclusions are set out at panel A of
Exhibit 13 appended to his report.
[115]
The RAROC analysis carried
out by Dr. Saunders calculated how much capital a third party would
require to support capital at risk of the Appellant, viewing the Appellant as
either AAA or AA. He used three models to calculate this. The average of
the three models after tax (assuming a tax rate of 40%) is 113% which is high
compared to the hurdle rate of 22%-23%,
that is, the RAROC is more than four times greater than the return on equity of
the stockholder, which suggests that the fee charged by GECUS was too high. He testified
that a reasonable guarantee fee in the circumstances would have been 20 basis
points.
[116]
Using Dr. Altman’s
Z-score double prime model, which he claimed is very relevant for determining
the creditworthiness of financial institutions, Dr. Saunders determined that
the Appellant’s creditworthiness improved during the years 1998‑2000. The
Appellant’s creditworthiness improved on the basis that its ability to pay debt
improved with the conversion of short‑term debt to long-term debt from a
related funding company in 1999 and 2000.
This result flows from the notion that investors are worried about ability to
pay debt, which he says is calculated by measuring current assets relative to
current liabilities. In his rebuttal report, Mr. Fidelman recalculated the
X1 variable in the Altman Z‑score double prime model to take into account
the misclassification of receivables as long term.
Dr. Saunders believed cutting down on commercial paper (i.e., short-term
debt) correlatively reduces current liabilities. He added that Mr. Fidelman
used the RiskCalc model that is built around manufacturing firms. In his view, the
three benefits to using the Z-score double prime model are that it was
calibrated just prior to the period relevant to this appeal, it is a private
sector model, and it focuses on non-manufacturing firms.
[117]
Dr. Saunders
testified to the effect that the only impact of GECUS dropping its explicit
guarantee with respect to the Appellant’s debt offerings would be a lowering of
the latter’s rating from AAA to AA. In his opinion, the U.S. parent would still have incentive to support the
Appellant in order to maintain its own AAA status, in the absence of an
explicit guarantee. According to Dr. Saunders, economic incentive is the
most important factor in finance.
[118]
Dr. Saunders contended
that the explicit written guarantee was a costly and unnecessary mechanism
given the strength of the economic incentive for the parent to provide the
Appellant with implicit support, but he agreed that the explicit guarantee achieved
the goal of securing a AAA rating for the Appellant. The
institutional investor, seeing the Appellant’s unguaranteed debt offerings, but
knowing they are issued out of Connecticut, by the same treasurer, by the same
company, and with the same liability structure would reason that it is not a
separate company issuing the debt.
[119]
In Dr. Saunders’
view, GECUS decided what capitalization level to impose on the Appellant. This
cannot be viewed as independent capitalization. Therefore, the actual
capitalization level of the Appellant should be ignored for the purposes of
determining its stand-alone credit rating. From the witness’s standpoint, what
determines whether the Appellant constitutes a core subsidiary under the
S&P criteria is the consolidated capital ratio, as opposed to the
independent capital ratio of the Appellant, since the Appellant did not choose
its own capital ratio. Further, given that the parent is AAA, it could inject
capital into the subsidiary at any time.
[120]
In relation to the
S&P criteria for classification as a core subsidiary, one of which is the inconceivability
that the unit could be sold when its “dependence upon the rest of the group
make[s] it impossible to sever the entity from the rest of the parent group”, Dr. Saunders
was of the view that, given the hub-and-spoke system and the push to maximize
geographic diversification, it is inconceivable that the Appellant would be
sold.
[121]
Dr. Saunders recognized
that GE Financial Assurance Holdings Inc. was rated A+, despite being a 100%-owned
subsidiary carrying the GE name and accounting for 20% of the assets of the
consolidated GECUS.
Testimony of Harold J. Meyerman
[122] Mr. Meyerman was qualified as an
expert in banking, not in rating matters.
He worked most of his professional career in the U.S.,
where he held very senior banking positions prior to his retirement. At the
initial stage of his career, for a very brief period, Mr. Meyerman was the
manager of a Canadian bank branch located in British Columbia.
[123]
The witness testified
that banks do their own credit risk analysis, typically using a bottom-up
approach, analyzing the various entities and then arrive at a consolidated
statement.
Banks use credit rating agency analyses as a check on their own internal
ratings.
[124]
Between 1995 and 1998, stated
Mr. Meyerman, North
America was an integrated
commercial paper market. Canadian investors would invest in the commercial
paper of U.S. issuers and vice versa.
[125]
According to the
witness, GE’s reputation was second to none. Even among AAAs, it was in a league
of its own. GE, during that period, was the most admired company in the world.
Major investors wanted to acquire GE or GECUS debt offerings. In that era,
GE or GECUS could dictate their own terms in respect of covenants.
[126]
Discussions between
Mr. Meyerman’s bank and GECUS about the Appellant issuing commercial paper
in its own name without an explicit guarantee would necessarily involve the
rating agency people coming on board to equalize the ratings, or at least get
close to doing so. Investors would require this.
[127]
The witness appeared to
be acutely aware of the fact that the removal of GECUS’ guarantee would not be
treated as welcome news by the credit rating agencies called upon to rate the
Appellant’s unguaranteed debt. In cross‑examination, the witness left me
with the impression that a high investment grade AAA credit rating for the
Appellant would have been more feasible had its debt been unguaranteed from the
outset. The witness asserted he would want to have a very good understanding of
the reasoning underlying Mr. Werner’s hypothetical decision to stop
guaranteeing the Appellant’s debt. While the witness insisted it would still be
possible to issue AAA paper if GECUS showed signs of support, he acknowledged
there would be a cost differential between two AAA note issues, one guaranteed
and the other unguaranteed.
[128]
While Mr. Meyerman
did not agree that the explicit guarantee was the most efficient or
cost-effective way of enabling the Appellant to enter the capital markets on
the financial strength of its parent, GECUS, he did view the explicit guarantee
as a more certain means to encourage investors. However, even where there is an
explicit guarantee, investors will still ask questions concerning credit
quality, but to a lesser extent.
[129]
Mr. Meyerman suggested
that, in a hypothetical world, one possible reason for dropping the parent company
guarantee would be to reduce the cost of funding the Canadian company, thereby
enabling it to compete more effectively. He did not accept the proposition that
dropping the guarantee leads to an inescapable inference that GECUS has decided
it wants to keep its options open in terms of whether or not it will come to
the rescue of the Appellant.
Testimony of Edward Emmer
[130]
Mr. Emmer was
called as an expert in credit rating analysis. He spent virtually his entire
career with S&P, starting there in 1969 and retiring in 2008. The majority
of that time was spent with the ratings division where, between 1992 and 2005,
he was head of global corporate and government ratings at S&P. The witness
was qualified as an expert in credit risk analysis relating to a broad cross‑section
of multinational entities, including industrial and global financial
institutions.
[131]
During his period at
S&P, it was common for Mr. Emmer to meet with GE executives. For
example, he attended, with fellow S&P employees, GE’s annual reviews, at
which GE executives were in attendance.
[132]
While his direct
involvement with GE executives decreased as his management responsibilities
grew in scope, they still met once a year. He was also kept informed by his
analysts. Additionally, if something major was transpiring which could have an impact
on GECUS’ AAA rating, he would participate in the discussions.
[133]
His assignment for the
purposes of the present case was to determine how S&P would rate the
unguaranteed debt of the Appellant. His conclusion was that the unguaranteed
debt of the Appellant would be rated the same as its parent company, GECUS.
[134]
He conceded having made
a computational error in relation to the size of the Appellant relative to
GECUS, as pointed out by Dr. Chambers. This error occurred from failing to
take into account the differences in currency values. Mr. Emmer added that
this computational error did not alter his overall opinion with respect to the
amount of support the parent would provide to its Canadian subsidiary: GECUS would
support the Appellant even if it withdrew its guarantee. The Appellant was
essentially an extension of the global business units.
[135]
Mr. Emmer referred
to two different approaches to determining ratings: the model-driven and analysis-driven
approaches. The model-driven approach works by analyzing financial ratios and
then putting them through a model, which in turn determines the rating.
Institutions with high amounts of loans outstanding tended to use this
approach. The analysis‑driven approach looks at many of the same factors
that are used as inputs into the model, but, in addition, this approach
considers certain qualitative, forward-looking factors, for example: what
management’s intentions and strategy were for their company.
[136]
S&P prefers the
analysis-driven approach, depending upon the situation, because the qualitative
factors can far outweigh the quantitative. In the present case, for example,
numbers express one thing, but on considering the ownership structure, namely the
fact that the Appellant is owned by GECUS, one may reach a far different
conclusion.
[137]
Mr. Emmer
conducted a credit rating pyramid analysis in connection with the Appellant. At
the relevant time, he found, the Appellant, because of its ownership by GECUS,
would be rated higher than the Canadian government’s external obligations. The quality of
GE management was excellent, which had a downstream effect on the management
decisions of the Appellant.
GE had a strong credit culture, which encompassed several characteristics: a
commitment to excellence, a sound value system, awareness of every
transaction’s impact on the bank, balancing the short and the long term,
respect for credit basics and no tolerance for surprises. These
characteristics would be overriding factors in determining how one would expect
a company to act in the marketplace with respect to its obligations.
[138]
In his appraisal of the
Appellant’s profile, as opposed to its credit rating, Mr. Emmer concluded
the profile would not suggest a AAA rating on a stand‑alone basis.
[139]
The fact that GECUS
guaranteed debt of the Appellant would be viewed by S&P as indicating that
the Appellant was an important subsidiary because otherwise GECUS would not
have guaranteed the debt. In Mr. Emmer’s view, GECUS would only guarantee
something they firmly stood behind and viewed as a long‑term, important
part of the company.
However, I noted during his testimony that he had difficulty explaining how the
Appellant and GECUS could portray a decision by GECUS to stop guaranteeing the
Appellant’s debt. If guaranteeing the debt meant GECUS would stand behind the
Appellant, did removing the guarantee signify the opposite? Mr. Emmer left
me with the impression that GECUS would not contemplate such a move because it would
disrupt the Appellant’s ability to raise capital quickly and cheaply.
[140]
This being said, the
witness maintained that S&P would view the Appellant as strategically
important to its parent despite its relatively low proportion of assets on a
consolidated basis. The Appellant has been doing business in Canada since 1963. Canada is neighbour to the U.S. and global expansion was an important strategic
objective of both GE and GECUS. For those reasons, the Appellant was an
important strategic asset that its parent would not walk away from.
[141]
In relation to
parent-subsidiary rating links, Mr. Emmer believed that parent companies
are going to do what is in their best interest. In the present case, it is
appropriate to equalize the ratings of the subsidiary and the parent. In the
case of entities that are dependent on financing in the capital markets, the
parent is less likely to let the subsidiary go.
[142]
Mr. Emmer advocated
a top-down approach to arrive at the Appellant’s rating. He viewed this as most
appropriate in this situation because he felt that the ownership of the
Appellant by GECUS was the most important factor. The unguaranteed debt of the
Appellant would have to be approved by its board of directors, which is
composed of senior executives from GECUS. The debt must also be approved by
GECUS to ensure compliance with its internal policies. The witness believed
that if the Appellant’s board and GE’s vice-president of finance authorized the
issuance of unguaranteed debt of the Appellant, they would stand behind that
debt, particularly considering that the letters GE are attached to the debt.
According to the witness, there is no way GECUS could walk away from the Appellant’s
debt in the capital markets, as opposed to debt of the Appellant resulting from
a private transaction, given that GECUS was dependent on the capital markets
for financing — it had to roll over $2 to $3 billion of
debt each day and it issued tens of billions of dollars of debt every year.
[143]
Mr. Emmer believed
Dr. Chambers placed too much emphasis on the as‑reported financial
statements of the Appellant and that he did not give enough weight to the fact
that the Appellant was owned by GECUS and ultimately by GE. Dr. Chambers also
underestimated the impact that a default of the Appellant would have, not only
on the Appellant, but on GECUS.
[144]
Mr. Emmer believed
the Appellant’s financial statements do not tell the whole story. He pointed
out that more time had probably been spent looking at the financial statements
of the Appellant in the past month than GE probably did in a whole five-year
period. In his view, the Appellant’s financial statements were merely an
accounting consolidation.
[145]
Mr. Emmer could
not reconcile his opinion with that produced by Dr. Chambers (AAA vs. BB),
nor did he believe a ratings committee would have accepted Dr. Chambers’
recommendation during the period in question. Dr. Chambers’
view was, in Mr. Emmer’s opinion, not within the realm of what could be
considered reasonable, given that Mr. Emmer was not aware of any situation
where a AAA-rated financial institution had a subsidiary carrying its name that
was rated BB.
[146]
According to
Mr. Emmer, the stand-alone analysis emphasized by Dr. Chambers is
circular and is an exercise in futility given that GECUS can make the Appellant
look, financially speaking, any way it wishes with no impact upon GECUS’
consolidated financial statements.
Therefore, he believed Dr. Chambers placed too much emphasis on the
Appellant’s capitalization ratio in his analysis.
[147]
In the absence of an explicit
guarantee, Mr. Emmer was not sure whether the parent company would have to
do something (beyond its implicit support) to enhance the balance sheet of the subsidiary
in order to make it match what the capital markets required under a stand-alone
analysis. In this sense, according to Mr. Emmer, the support is explicit
as opposed to implicit.
However, assuming intact ownership, and everything else being equal, he believed
this would result in the rating recommendation he said he would make to the ratings
committee.
[148]
In a worst case, yet
reasonable, scenario in which the Appellant was allowed to default, GECUS would
face difficulty in rolling over, or become unable to roll over, its maturing
debt obligations. Therefore, if the market began to lose confidence in GECUS as
a result of the Appellant’s default, it would have about one week to improve
market sentiment because even if it completely drew down its bank lines it
would not have sufficient liquidity to repay the commercial paper. As a result, Mr. Emmer
affirmed, GECUS would do anything to prevent a loss of market confidence in it.
[149]
Finally, he agreed that
there was clearly a business reason for the guarantee: it provided a
straightforward mechanism for ratings equalization. However, according to the
witness, this does not mean that the Appellant would not have been AAA without
the guarantee.
Testimony of Brian Becker, Ph.D.
[150]
Dr. Becker holds a
Ph.D. in applied economics from the Wharton School of the
University of Pennsylvania.
Dr. Becker was qualified as an economist specializing in transfer pricing.
[151]
The witness testified
that an arm’s length valuation involves a two-step approach. The first is to
define what it is you are valuing, otherwise, it is difficult to ascertain what
evidence to look for in trying to arrive at the value. In this case,
determining how different strategic factors and characteristics would fit into
an arm’s length valuation is not simple. The second step is to look for
evidence of what the price might be in the marketplace directly, through
prices, profit margins, ratings and anything that may be of value. Then, one
must sort through all of this information to find that which is more useful, making
any adjustments that may be necessary here and there.
[152]
In step one, Dr. Becker
described the characteristics of the relevant transaction. He then proceeded to
determine what the transaction would look like if the parties were at arm’s length. According to
the witness, one of the important characteristics in this case is the
organizational structure of the parties.
[153]
In defining the
transaction, regard should be given not only to the terms of the transaction,
but also to the behaviour of the parties. The risks of the parties in the
hypothetical debtor and hypothetical guarantor transaction should match that of
the parties in the actual transaction.
[154]
In his rebuttal report,
Dr. Becker discussed the imprecision of the construct adopted by Mr. Cole
who, in his report, maintained the guarantor in the same structure/position,
but recast the position of the party receiving the guarantee as a stand-alone
company. By making a downward adjustment for only one of the parties to the
transaction, Mr. Cole rendered his construct biased and imprecise.
[155]
Transfer pricing is to
be approached from both sides, from the guarantor’s side in terms of what
its costs are and from the debtor/beneficiary’s side in terms of the benefit to
it. Both parties must be satisfied in order for them to agree to a transaction.
Dr. Becker took issue with Mr. Fidelman’s insurance‑based model
on the basis that its application considered the issue only from the standpoint
of the guarantor. Looking only at the cost in some calculation on one side does
not, unless the other side is willing to pay that cost, lead to the completion
of a transaction.
[156]
Mr. Fidelman’s
model, by working from the point of view of costs to the guarantor, does not take
into account what the debtor might be willing to pay given the existence of
implicit support. In the witness’s view, taking that aspect into account would
in effect lead to the determination of a lower fee that the debtor would be
willing to pay for the guarantee.
[157]
Counsel for the
Respondent referred to Mr. Cole’s testimony to the effect that account
should not be taken of the parent-subsidiary relationship in an arm’s length
valuation. In response to this, Dr. Becker remarked that Mr. Cole
actually was taking into account the parent-subsidiary relationship, at least
with respect to the guarantor, by refusing to view the guarantor as a stand-alone
entity. Further, the parent-subsidiary relationship and the entire
organizational structure does potentially affect how one defines and
characterizes the transaction, and in this regard, the transfer pricing Dr. Becker
was familiar with already incorporates the parent-subsidiary relationship
(i.e., the organizational structure).
Testimony of Deloris Wright, Ph.D.
[158]
Dr. Wright received
a bachelor of science degree in business from Oklahoma State
University and a Ph.D. in economics from Iowa
State University.
She was qualified as an expert economist specializing in transfer pricing. Dr. Wright
was not asked to do a transfer pricing study. Rather she was asked to critique
the Appellant’s analysis.
[159]
According to the
witness, the first step in the analytical approach to determining arm’s length
prices is to identify the transaction under review. In so doing, she assumed that
the parties were unrelated but every other aspect of that relationship remained
unchanged.
She would look at both sides of the transaction. In respect of both parties, the
question to ask is: what are the functions performed and the risks incurred?
The next step involves asking oneself the meaning of these facts and then considering
whether there is a comparable market transaction in which unrelated parties
deal with each other in the same way as the Appellant and GECUS.
[160]
Noting there were no perfect
comparables, Dr. Wright stated that the next step is to evaluate the
comparable transaction in relation to the relevant one and to ask: what are the
differences between the comparable and the relevant transaction that have an impact
on the price that unrelated parties would agree to? Adjustments are then made
for these differences. At this point, the result will be a reliable arm’s
length price,
but the comparable may be rejected from the outset if it is too different. The parties
can resort to a different method in that case.
[161]
Two additional facts
came to Dr. Wright’s attention when reading the trial transcripts. First,
nobody from the Appellant’s side participated in the transaction. Second,
because there were 50-60 players in the commercial paper market, GECUS was
issuing its debt to a small, integrated global capital market. Thus,
reputational issues become very important.
[162]
The witness asserted
there is no comparable transaction here because a third‑party guarantor
would never manage the risk as GECUS does in the present case. Without a
comparable, the manner in which the witness analyzed the relevant transaction
is as follows: all the decisions are made in the U.S.,
so what price would the U.S. parent charge and what price would the Canadian
subsidiary be willing to pay? The Appellant would say that the U.S. parent is making all of the decisions and controlling
all the risk. This ensures that the Canadian subsidiary (the Appellant) will
not default on the loan.
[163]
The witness did,
however, acknowledge that GECUS takes on risk as a result of the guarantee
(e.g., the risk of a flight to quality government paper only), even though it
has supervisory control over the subsidiary.
There is also a benefit to the Appellant in that GECUS cannot abandon it in the
event of a liquidity crisis or a flight to quality that renders it
prohibitively expensive or impossible to refinance the debt.
[164]
Dr. Wright stated
that a number of factors must be considered in determining an arm’s length
price for the guarantee. First, account should be taken of the parent’s
reputation (i.e. implicit support). To the extent that the analysis has already
taken this into account, no further adjustment should be made. Additionally, no
covenant was required from the Appellant as a result of the guarantee, which is
a benefit to the Appellant. This should be reflected in the amount of the fee.
It has the impact of increasing the fee. Next, there were no reserves for loss
contingencies in GECUS, but Mr. Fidelman computed his fee by assuming the
insurer would have a loss contingency reserve and there would need to be a
return on that. This would need to come out of the fee. Lastly, regard should
be given to what the witness referred to as the substance of the transaction,
the question being in an unrelated-party situation, would the debtor pay a fee
to the guarantor where the guarantor has complete control over the default risk? I note that
the witness also agreed with the Court’s observation that if the guarantee is
withdrawn, value should also be attributed to the negative impact that would
have on the market’s perception of the parent’s willingness to support the
Appellant.
Dr. Wright also agreed with the position that benefits accruing to GECUS
qua shareholder should not be taken into account in the determination of a market
price for the transaction.
[165]
According to the
witness, Mr. Cole’s analysis did not make a comparison between the
intercompany facts and the third-party facts. Dr. Wright believed that
Mr. Cole’s results were unreliable because he failed to consider adjustments
in his analysis.
[166]
Dr. Wright believed
GECUS would have been willing to charge the Appellant a low guarantee fee
because the potential cost to GECUS of a default by the Appellant would have
been much greater than the potential benefit that it would have got from an
arm’s length guarantee fee.
POSITION OF THE PARTIES
Respondent’s Position
[167]
Counsel for the
Respondent posits that former subsection 69(2) and current paragraph 247(2)(a),
which are the relevant provisions of the Income Tax Act (the “ITA”)
applicable to the issue herein, conform with the principles enunciated in the
Organisation for Economic Co‑operation and Development guidelines (the “OECD
Guidelines”) on transfer pricing.
In this regard, chapter 7 of the OECD Guidelines provides specific commentary
on intra-group services. This chapter informs us on two key issues: whether an
intra-group service has, in fact, been provided and, if so, what the proper
arm’s length price is for the service. One of the key principles relied on by
the Respondent in this case is enunciated at paragraph 7.13, which reads as
follows:
Similarly, an associated enterprise should not be considered to
receive an intra‑group service when it obtains incidental benefits
attributable solely to its being part of a larger concern, and not to any
specific activity being performed. For example, no service would be received
where an associated enterprise by reason of affiliation alone has a credit rating
higher than it would if it were unaffiliated, but an intra-group service would
usually exist where the higher credit rating were due to a guarantee by another
group member, or where the enterprise benefited from the group’s reputation
deriving from global marketing and public relation campaigns. In this respect,
passive association should be distinguished from active promotion of the MNE
group’s attributes that positively enhances the profit‑making potential
of particular members of the group. Each case must be determined according to
its own facts and circumstances.
[Emphasis added by the Respondent.]
[168]
The Respondent submits that
the Appellant’s credit rating would be equalized with that of GECUS by reason
of affiliation in the absence of a guarantee arrangement. On this theory, the
Respondent claims that the Appellant could have borrowed the same amount of
money at the same interest rate without an explicit guarantee as it did with
such a guarantee. As a result, the Appellant did not receive an economic
benefit from the guarantee. In this case, the arm’s length price for the
guarantee is nil. The guarantee arrangement was simply a clearer indication of
the implicit support that already existed in favour of the Appellant.
[169]
Counsel invites me to
consider the credit rating methodology developed and applied by S&P in the
taxation years under review in assessing whether or not the Appellant’s credit
rating would be equalized with that of its parent in the absence of an explicit
guarantee. Under S&P’s credit rating system, the Respondent claims, the
Appellant would be considered a “core subsidiary”. According to the Respondent,
the crucial point is that the Appellant’s credit rating would be notched up to
the AAA rating of its parent, GECUS, on the grounds that both S&P and the
Appellant’s debt holders would recognize that GECUS had a strong economic
incentive to provide financial support to the Appellant in times of financial
stress, even if it was not contractually obliged to do so.
[170]
This economic incentive
originates from GECUS’ AAA credit rating, which allows it to borrow large
amounts of money at very low interest rates. This highest investment grade
rating is a significant comparative advantage as few enterprises active in the
unregulated financial sector are rated AAA. The Respondent argues that GECUS
would suffer a significant credit rating downgrade if a like-named affiliate,
such as the Appellant, was allowed to default on its debt. In such a case, the
additional borrowing costs to GECUS would exceed the cost of financial support
to the Appellant, thus prompting GECUS to intervene. Therefore, GECUS would not
sacrifice its AAA rating by allowing the Appellant to default on unguaranteed
debt.
[171]
Counsel for the
Respondent points out that Mr. Emmer, who was qualified as an expert on credit
rating methodologies, testified that this type of “implicit support” would be
recognized by S&P in the case of the Appellant and that its credit rating
would be equalized with that of GECUS even in the absence of an explicit
guarantee.
[172]
As an alternative
position, the Respondent invites me to adopt a two‑step approach in the
event that I find the Appellant’s credit would not be equalized with that of GECUS.
As a first step, I should identify the interest rate spread based on my
conclusions regarding the difference between the Appellant’s and GECUS’ credit
rating (the “yield approach”). As a second step, appropriate adjustments should
be made to the interest rate differential to take into account the benefits
flowing to GECUS under the guarantee arrangement. The Respondent claims these
benefits would be taken into account in arm’s length bargaining.
Appellant’s Position
[173]
The Appellant advances
two legal arguments in opposition to the Crown’s approach to this case.
Mr. Meghji, counsel for the Appellant, points out that the Crown’s
approach presupposes that the benefits that flow from the parties’ relationship
should be taken into account in determining whether the 100-basis point annual
fee paid by the Appellant for the guarantee exceeds an arm’s length price.
Counsel argues the affiliation benefit, if any, enjoyed by the Appellant from
its place in the GECUS organizational structure cannot be considered under
subsection 69(2) and paragraph 247(2)(a) of the ITA. The scheme
laid out by these provisions requires that one situate the parties opposite each
other to determine how they would have arranged their transaction if they had
been dealing at arm’s length. All distortions that arise from the parties’
relationship must be eliminated to arrive at an arm’s length result.
[174]
If an affiliation
benefit exists, as alleged by the Crown, it exists because GECUS indirectly
owns all of the shares of, and controls, the Appellant. This factor is not
present in an arm’s length relationship and must be discarded to determine an
arm’s length price for the transaction. It follows from this premise that the
Appellant’s credit rating prior to the implementation of the explicit guarantee
must be determined solely on a stand-alone basis without factoring in any
credit rating for the implicit support of GECUS. On a stand-alone basis,
counsel contends, the evidence shows the Appellant’s credit rating to be, at
best, BB in the years in question. As a result, under the yield approach
proposed by the Crown, the economic benefit enjoyed by the Appellant under the
guarantee arrangement far exceeds the fee paid to GECUS.
[175]
Counsel also argues that
the factors which the Crown invites me to consider and adjust for under its
alternative two-step approach are benefits that accrue to GECUS qua
shareholder. For example, the prospect of higher dividends due to enhanced
profits derived from interest cost savings accrue to GECUS because of share
ownership and not by virtue of the guarantee arrangement. This benefit is
attributable to the fact that GECUS assumed the risk of investing capital in
the business of the Appellant. All benefits that are attributable to share
ownership should be ignored.
[176]
Finally, the Appellant submits
that the Crown’s approach in this case is flawed because it is based on the proposition
that the explicit guarantee was not needed by the Appellant in the
circumstances. Counsel posits that this is why the Crown led evidence to
convince me that the Appellant would be rated AAA even without an explicit
guarantee from its parent. According to the Appellant, subsection 69(2) and
paragraph 247(2)(a) of the ITA require that the transaction
arranged by the parties be priced by the Minister. The parties entered into a
guarantee transaction, and the yield approach put forward by the Minister merely
determines the cost at which the Appellant can borrow money, it does not seek
to address the question of what a guarantee would cost in similar
circumstances. The Minister did not bother to ask himself whether an arm’s
length party would have provided a guarantee to the Appellant because the
Minister had decided that the guarantee was simply not required in the
circumstances. For the Respondent to succeed, the reassessments would have had
to have been based on paragraph 18(1)(a) of the ITA.
[177]
As a final point, the
Appellant submits that the Crown’s theory in this case is not supported by the
evidence adduced at trial. In brief, the Appellant argues that the evidence demonstrates
the following:
(a) The explicit guarantee
was implemented to satisfy bona fide business needs of the Appellant. This is
confirmed by the fact than an explicit guarantee has been in place from 1988 to
the present day. It was only in 1995 that a guarantee fee was charged. The
Respondent now accepts this fact as it has dropped its reliance on paragraph
247(2)(b) of the ITA.
(b) The explicit guarantee
benefited the Appellant’s lenders, who obtained the right, in the event of the
Appellant’s default, to demand payment from GECUS, which became legally bound
to discharge the Appellant’s obligations as a result of the explicit guarantee.
(c) Even if one were to
accept the Crown’s theory regarding the case, the Appellant would not have
obtained a AAA rating from S&P, or any other rating agency for that matter,
if implied support was taken into account in the analysis. The Appellant’s
credit rating would have been between BB and BB‑ without an explicit
guarantee.
(d) The rating agencies
and investment dealers would demand that the Appellant maintain standby letters
of credit in the amount of $3 billion if it tried to gain access to the
Canadian commercial paper markets without an explicit guarantee. The Appellant
could not negotiate standby credit facilities without the explicit support of
its parent.
ANALYSIS
Legal Framework
[178] I will examine the competing legal
positions presented by the parties before turning my attention to the evidence
in this case.
Is a Stand-Alone Credit Rating the Proper Analytical Approach?
[179]
The parties argue that the
wording differences between subsection 69(2) and paragraph 247(2)(a) of
the ITA are not meaningful for the purpose of this appeal. In addition,
both parties agree that it is the transaction arranged by the parties that is
the object of the transfer pricing analysis. The parties disagree on which of
the economically relevant characteristics of the guarantee arrangement should
be considered by me in determining an arm’s length price for the guarantee
transaction.
[180]
The Appellant maintains
that the Respondent’s approach is flawed because subsection 69(2) and paragraph
247(2)(a) require me to situate the Appellant and GECUS opposite each
other as arm’s length parties in order to ascertain whether the price agreed to
meets an arm’s length standard. According to the Respondent, the words of
subsection 69(2) require that I ascertain whether the amount paid was “. . .
greater than the amount . . . that would have been reasonable in the
circumstances if the non‑resident person and the taxpayer had been
dealing at arm’s length . . .”. The concept of “implicit support”
relied on by the Crown to convince me that the Appellant’s credit rating would
be equalized with that of GECUS requires that one preserve the very non‑arm’s
length relationship which subsection 69(2) and paragraph 247(2)(a)
invite me to ignore. Stated differently, the reputational pressures that may
cause GECUS to support the Appellant in times of financial stress exist because
the Appellant is allegedly a core subsidiary. This type of pressure does not
exist in an arm’s length relationship. All factors of influence flowing from
the non-arm’s length relationship must be ignored to ensure an arm’s length
result. The risk assumed by one party to a transaction has no bearing on the
default risk of the other party prior to the implementation of the guarantee
arrangement. In the present case, the failure of the Appellant to meet its
financial obligations would taint its parent’s reputation because AAA issuers
are expected not to allow their core subsidiaries to default.
[181]
I believe a careful
analysis of the scheme of the transfer pricing rules is required to dispose of
this issue. Because the parties agree there is no meaningful distinction
between the scope of subsection 69(2) and paragraph 247(2)(a) of the ITA,
I will deal with the text of the latter provision for the reasons that it is
more complete and that it represents Parliament’s last word on the matter.
[182]
Paragraph 247(2)(a)
of the ITA provides that the Canadian transfer pricing rules apply to
parties that do not deal with each other at arm’s length. For the purposes of
the ITA, related persons are deemed not to deal with each other at arm’s
length.
In a corporate context, paragraph 251(2)(b) of the ITA states that
corporations are related to:
(i) a person who controls the corporation, if it is controlled by
one person,
(ii) a person who is a member of a related group that controls the
corporation, or
(iii) any person related to a person described in subparagraph
251(2)(b)(i) or 251(2)(b)(ii) . . . .
[183]
Similarly, paragraph
251(2)(c) provides that two corporations will be considered related:
(i) if they are controlled by the same person or group of persons,
(ii) if each of the corporations is controlled by one person and the
person who controls one of the corporations is related to the person who
controls the other corporation,
(iii) if one of the corporations is controlled by one person and
that person is related to any member of a related group that controls the other
corporation,
(iv) if one of the corporations is controlled by one person and that
person is related to each member of an unrelated group that controls the other
corporation,
(v) if any member of a related group that controls one of the
corporations is related to each member of an unrelated group that controls the
other corporation, or
(vi) if each member of an unrelated group that controls one of the
corporations is related to at least one member of an unrelated group that
controls the other corporation.
[184]
The common thread in
each of the above cases is de jure control. De jure control means
the right of control residing in the ownership of the shares which carry the
majority of the voting rights that can be exercised to elect the majority of
directors to a corporation’s board.
[185]
Paragraph 251(1)(c)
provides that it is a question of fact whether unrelated parties are dealing with
each other at arm’s length. The courts have developed the following criteria to
determine this factual question:
(a) Was there a common
mind directing the bargaining of both parties to the transaction?
(b) Were the parties to the
transaction acting in concert, without any separate interest?
(c) Was there de facto
control?
The common thread here is de facto control.
[186]
The parties do not
dispute that GECUS and the Appellant are related and not dealing at “arm’s
length” by virtue of the de jure control that the former entity has over
the latter.
[187]
Paragraph 247(2)(a)
also indicates that the “arm’s length” principle must be applied by the trier
of fact in determining whether an adjustment of an amount otherwise determined
for the purposes of the ITA is required to match the “quantum or nature”
of the amounts that would have been determined if the terms and conditions of
the transaction had been made by arm’s length parties. In this case, the crux
of the dispute is whether the guarantee fee paid by the Appellant to GECUS
exceeds an arm’s length price. There are two opposing positions that I must
choose between in determining this question. Do all of the economically
relevant factors have to be considered in the determination of an arm’s length
price for the transaction in order to arrive at a meaningful comparison, as
suggested by the Crown? Does the scheme of paragraphs 247(2)(a) and (c)
suggest that all factors which are particular to the non‑arm’s length
relationship must be discarded, as suggested by counsel for the Appellant? What
do the relevant transfer pricing provisions say on this point? Before answering
this question, it is important to recall the applicable principles of statutory
interpretation.
[188]
The words of a statute
cannot be altered by a textual, contextual, and purposive analysis when they
are clear and plain. One would simply interpret the words or expression according
to their ordinary meaning, which is often referred to as the plain meaning
rule. However, the expressions “arm’s length” and “non‑arm’s length” are
creations of law. They are not words of ordinary language from which a plain
meaning can be easily distilled. In this case, the parties debated how the
expression “arm’s length” should be interpreted in a hypothetical world, thereby
inevitably making the legal and practical effect of the transaction somewhat uncertain.
Starting from the premise that the meaning of the expression is not clear and
plain, one must turn to the textual, contextual and purposive analysis to
clarify the expression in the context of transfer pricing.
[189]
The modern approach to
statutory interpretation was stated by the Supreme Court of Canada in Canada
Trustco Mortgage Co. v. Canada:
10 It has been long established as a matter of statutory
interpretation that “the words of an Act are to be read in their entire context
and in their grammatical and ordinary sense harmoniously with the scheme of the
Act, the object of the Act, and the intention of Parliament”: see 65302
British Columbia Ltd. v. Canada, [1999] 3 S.C.R. 804, at para. 50. The
interpretation of a statutory provision must be made according to a textual,
contextual and purposive analysis to find a meaning that is harmonious with the
Act as a whole. . . .
[Emphasis added.]
[190]
The origin of the
modern rule is often attributed to E.A. Driedger, he who wrote that “the words
of an Act are to be read in their entire context and in their grammatical and
ordinary sense harmoniously with the scheme of the Act, the object of the Act,
and the intention of Parliament”.
[191]
As noted in Minister
of National Revenue v. Sheldon’s Engineering Ltd., when
considering the phrase “dealing at arm’s length”:
. . . the Income Tax Act . . . does not
purport to define the meaning of the expression generally: it merely states
certain circumstances in which persons are deemed not to deal with each other
at arm’s length.
[192]
While a number of
cases, including many in a taxation context, have attempted to define the
phrase “non-arm’s length”, few have discussed the converse, namely the meaning of
the phrase “arm’s length”.
[193]
The meaning of arm’s
length was considered extensively in Crawford & Co. v. Canada by
Porter J. The provision being considered, subsection 5(2) of the Employment
Insurance Act,
defers to paragraph 251(1)(b) of the ITA for guidance on the
arm’s length issue. As noted by Porter J., the majority of cases
considering the phrase “arm’s length” in Canada
have “tended to involve one person (either legal or natural) controlling the
minds of both parties to the particular transaction”. More specifically, he
stated:
29 However, simply because these leading cases involved such
factual situations, does not mean that people who might ordinarily be in a non arm’s
length relationship cannot in fact “deal with each other at a particular time
in an ‘arm’s length’ manner”, any more than it means that people who are
ordinarily at arm’s length might not from time to time deal with each other in
a non arm’s length manner. These cases are quite simply examples of what is not
an arm’s length relationship rather than amounting to a definition in positive
terms as to what is an arm’s length transaction. Thus at the end of the day
all of the facts must be considered and all of the relevant criteria or tests
enunciated in the case law must be applied.
[194]
The meaning of arm’s
length was considered outside a tax context by the British Columbia Court of
Appeal in Re Galaxy Sports Inc.,
and that court defined the term as meaning “no bonds of dependence, control or
influence, between the corporation and the person in question”, in the sense
that the latter has no moral or psychological leverage sufficient to diminish or
possibly influence the free decision making of the former.
[195]
More consistent with
economic theory, is the definition adopted by the Court of Queen’s Bench of
Alberta in Pocklington Foods Inc. v. Alberta (Provincial Treasurer):
197 . . . In assuming that the parties are acting at
arm’s length, the negotiation is contemplated to be between parties with
opposing interests, each having an economic stake in the outcome.
[196]
With this background in
mind, the concept of “dealing at arm’s length” used in the context of the
transfer pricing rules to determine a market price for a transaction refers to
how independent parties negotiating with each other in the marketplace would
behave — the vendor or service provider, for the
purpose of achieving the highest price or best terms for his goods or services,
and the other party, the purchaser, for the purpose of acquiring the goods or
services at the lowest price.
[197]
In the final analysis,
the “arm’s length” principle in the transfer pricing context is tied to modern
economic theory, which is based on observations of how parties act in the
marketplace. Economic theory assumes that individuals in the marketplace will
employ a cost benefit analysis in choosing among the alternatives available for
achieving their commercial objectives. The arm’s length principle also embodies
other features of general human behaviour. Market actors will seek out all
relevant information, including information that helps them to understand their
counterpart’s motivation for entering into a transaction with them.
Dr. Becker advanced these concepts in his report and during his testimony
at trial.
[198]
The question becomes
one of fact or, more precisely, one of identifying the economically relevant
characteristics of the transaction that may influence the arm’s length parties
in their negotiations.
[199]
In my opinion, counsel
for the Appellant misapplied the arm’s length principle when he suggested to me
that the concept of “implicit support” should be ignored because it is rooted
in the non-arm’s length relationship. That concept has nothing to do with the
exercise of de facto or de jure control which defines a non‑arm’s
length relationship. The reputational pressure is exerted by GECUS’ debt
holders. It is GECUS’ debt holders that would react negatively if the Appellant
was allowed to default on its debt. The fact that both GE and GECUS prize their
AAA rating is commonly known in the marketplace. In many instances, it was GE
and GECUS that advertised the materiality of their AAA credit rating in their
public press releases and annual returns.
[200]
I can easily imagine how,
in the present day context, the iconic investor, Warren Buffet, would have
known that GE valued its AAA credit rating and that the loss of this status
would entail significant additional financial charges for all of its
operations. This factor would undoubtedly have been taken into account in
designing the terms and conditions of, and determining the dividend rate and
the conversion premium for, the GE preferred shares for which Berkshire
Hathaway ultimately subscribed.
Undoubtedly GE would have weighed the pros and cons of agreeing to a private
placement rather than issuing a public offering or lobbying the U.S. government for credit enhancement or support.
Successful investors and investees are undoubtedly always attempting to use all
information and current circumstances to their advantage.
[201]
The evidence which is
discussed later in this judgment shows that anyone contracting with GE or GECUS
would be aware of the fact that GE or GECUS would want to preserve its high
rating. Dr. Chambers, who appeared for the Appellant, accepts the fact that
implicit support could lead to a double notch increase in the Appellant’s
credit rating for unguaranteed debt over a stand-alone rating. The question of
fact then becomes how successful this negotiating approach would be, having
regard to all of the other relevant facts and circumstances, some obviously
countervailing, that would be taken into account in arm’s length negotiations.
[202]
My interpretation of
the “arm’s length” principle used in the context of subsection 247(2) is
consistent with the decision of my colleague, Rip A.C.J. (as he then was), in GlaxoSmithKline
Inc. v. The Queen.
In that case, he recognized that the arm’s length principle enunciated in
the OECD Guidelines inform us as to the proper interpretation and application
of former subsection 69(2), as follows:
59 Subsection 69(2) of the Act is analogous to Article 9(1) of
the OECD Model Double Taxation Convention on Income and Capital. The
OECD issued a commentary on transfer pricing analysis in 1979. The Canada
Revenue Agency ("CRA") relies on OECD Commentaries in assessing:
Information Circular 87-2, International Transfer Pricing and Other
International Transactions, dated February 2, 1987. Information Circular
87-2 was replaced by IC 87-2R, International Transfer Planning on
September 27, 1999. The Federal Court of Appeal has said that it is
"common ground that the [OECD Commentary] inform or should inform the
interpretation and application of subsection 69(2)".
60 The OECD Commentary on Article 9(1) relies on the arm's
length principle to determine the prices that multinational enterprises
("MNEs") would charge for goods and services sold from one
jurisdiction to another. The arm's length principle recognizes that independent
enterprises would charge prices according to market forces when dealing with
each other. The Commentary recognizes that transfers between MNEs do not
necessarily represent the result of free market forces, but may instead have
been adopted for the convenience of the MNE. Consequently, prices set by an MNE
may differ significantly from the prices agreed upon between unrelated parties
engaged in the same or similar transactions under the same or similar
conditions.
[203]
Rip A.C.J. was applying
principles accepted by the Federal Court of Appeal in SmithKline Beecham
Animal Health Inc. v. Canada.
[204]
Paragraph 1.6 of the
OECD Guidelines states that the concept of independent parties is used to
adjust profits “by reference to the conditions which would have been obtained
between independent enterprises in comparable transactions in comparable
circumstances”.
Paragraph 1.15 of the OECD Guidelines reinforces this principle by providing as
follows:
Application of the arm’s length principle is generally based on a
comparison of the conditions in a controlled transaction with the conditions in
transactions between independent enterprises. In order for such comparisons
to be useful, the economically relevant characteristics of the situations being
compared must be sufficiently comparable. To be comparable means that none of
the differences (if any) between the situations being compared could materially
affect the condition being examined in the methodology (eg. price or margin),
or that reasonably accurate adjustments can be made to eliminate the effect of
any such differences . . . . This point is relevant to the question
of comparability, since independent enterprises would generally take into
account any economically relevant differences between the options realistically
available to them (such as differences in the level of risk or other
comparability factors discussed below) when valuing those options.
[Emphasis added.]
[205]
This is a clear
articulation of the importance of maintaining the relevant economic
characteristics of the controlled transaction in order to ensure the
reliability of the comparisons with uncontrolled transactions.
Is Reliance on Paragraph 18(1)(a) Required in Order to
Uphold the Reassessments?
[206]
Counsel for the
Appellant also takes issue with the Respondent’s approach in this case on the basis
that the Minister failed to ask a critical question required under subsection 69(2) and paragraph 247(2)(a) of the ITA.
In the Appellant’s view, the Crown has built its entire case on the grounds
that the explicit guarantee was not required in the circumstances because the
Appellant’s credit rating would have been equalized with that of its parent in
any event. According to the Appellant, subsection 69(2) and paragraph 247(2)(a)
of the ITA require the Minister to inquire how non-arm’s length parties
would have priced the transaction in similar circumstances. This was not done.
The Crown failed to adduce pricing evidence at trial as required by the
relevant provisions.
[207]
I believe counsel has
oversimplified the Crown’s approach to this case. In fact, the Respondent did
not argue that the guarantee fee was not incurred for the purpose of earning
income, which is the test prescribed by paragraph 18(1)(a). The starting
point for the Crown’s approach is the creditworthiness of the Appellant before
the explicit guarantee was implemented. The Respondent argues that the object
of a guarantee is credit enhancement. To determine whether or not, in fact, the
explicit guarantee mitigated the default risk of the Appellant’s debt
offerings, one must first determine the Appellant’s credit rating without the
explicit guarantee.
[208]
According to the Respondent,
because the guarantee does not enhance the Appellant’s credit, the Appellant
would not have entered into the guarantee contract with GECUS if the parties had
been unrelated. The Respondent contends that a AAA-rated issuer would not enter
into a guarantee contract in similar circumstances.
[209]
Stated differently,
because of implicit support there would be a ratings equalization between the
Appellant and its stronger parent, GECUS. In that case, the Appellant would
have no reason to seek out a third-party guarantor in the marketplace. There
could be someone willing to provide the guarantee for a price higher than zero,
but the Appellant would be unwilling to buy the guarantee if it provided no
economic benefit. Therefore, it is not surprising that no market data can be found
that can be used to price the transaction.
[210]
I share this point of
view. I believe an arm’s length person would not contract for a service if that
person feels the service would provide no benefit in the circumstances. I find
the above argument to be persuasive on this point, provided that the evidence considered
later on in this judgment demonstrates that the Appellant received no benefit, or
little benefit, from the guarantee arrangement.
[211]
I note that the case
law on a provision which is similar to subsection 69(2) and paragraph 247(2)(a)
does not support the Appellant’s contention that neither of these provisions
can be used to eliminate the expense in its entirety. The case law has held
that section 67 of the ITA can be relied on by the Minister to disallow
an expense, in whole or in part, if it can be shown that the expense was
unreasonable in the circumstances. Section 67 operates much in the same way as
subsection 69(2). The latter provision focuses on the reasonableness of an
expense incurred by a taxpayer in a cross-border related-party transaction.
[212]
This point was
considered by the Federal Court of Appeal in Hammill v. Canada, a case in
which that court dismissed the appellant’s argument that section 67 could not
be used to reduce an expense to zero, as follows:
52 In devising the “recommended approach”, the Supreme Court
identified section 67 as the statutory means of controlling excessive or
unwarranted expenditures once a source of income is found to exist. It said at
paragraph 57:
... If the deductibility of a particular
expense is in question, then it is not the existence of a source of income
which ought to be questioned, but the relationship between that expense and the
source to which it is purported to relate. The fact that an expense is found to
be a personal or living expense does not affect the characterization of the
source of income to which the taxpayer attempts to allocate the expense, it
simply means that the expense cannot be attributed to the source of income in
question. As well, if, in the circumstances, the expense is unreasonable in
relation to the source of income, then s. 67 of the Act provides a
mechanism to reduce or eliminate the amount of the expense. Again,
however, excessive or unreasonable expenses have no bearing on the
characterization of a particular activity as a source of income. [emphasis
added]
53 The choice of words (reduce or eliminate) is not
accidental. The Supreme Court was setting-up section 67 as the proper means of
testing the reasonableness of an expense once a business has been found to
exist. It was doing so after having explained that at the first level of
inquiry (i.e. the existence of a source of income and the relationship between
an expense and that source) courts ought not to second guess the business
judgment of the taxpayer (Stewart, supra, paragraphs 55, 56 and 57).
Section 67 was identified as the statutory authority pursuant to which an
inquiry could be made as to the reasonableness of an expense. In my view, the
Supreme Court in Stewart acknowledged that there is no inherent limit to
the application of section 67, and that in the appropriate circumstances, it
can be used to deny the whole of an expense, if it is shown to be unreasonable.
[213]
There is no reason why
I should depart from this principle in the instant case. During argument, I
asked counsel for the Appellant whether he agreed with me that his approach
meant that the Minister could not challenge a stewardship service or a duplicative
service under the transfer pricing rules relied on in the instant case. Counsel
for the Appellant agreed with this observation. I believe the principles
enunciated in the Hammill case prove that this argument is flawed. A
transfer pricing charge can be eliminated under the transfer pricing rules
without resorting to a claim made under paragraph 18(1)(a), provided that
it is established that the value of the benefit received from the service is
nil.
The Role of Expert Witnesses
[214] Before embarking on a detailed analysis of
the evidence and the technical opinions presented by the experts at trial, I
believe it would be useful to remind ourselves of the role an expert is called upon
to play and the approach a trial judge should take in hearing opinion evidence.
[215]
First and foremost,
with regard to expert evidence, I believe it is essential that the person
deciding the case fully understand the witness’s opinion before deciding to
accept it in whole or in part or to reject it. I surmise that counsel on both
sides spent a considerable amount of time coming to an understanding of the opinions
the experts were called upon to articulate at trial. Counsel for both sides
demonstrated they were extremely well prepared for the trial and had acquired a
mastery of highly technical financial information and theories. I complimented
both sides at the end of the trial for their thoroughness at the hearing.
Nonetheless, there were a number of technical points covered by the experts on
each side that required considerable clarification. For example, the Altman Z-score
double prime model used by Dr. Saunders in his report is highly technical
and is difficult for a lay person to understand at first glance.
[216]
Bryant, Lederman and
Fuerst write in The Law of Evidence in Canada:
The expert witness should provide independent assistance to the
court and should not assume the role of an advocate. An expert should state the
facts or assumptions upon which his or her opinion is based and should not omit
to consider material facts which weaken his or her position.
[217]
Similarly, Jean-Claude
Royer in La preuve civile describes the expert’s responsibility as
follows: “L’expert doit être impartial. Son rôle est d’éclairer le tribunal et
non d’être l’avocat d’une partie.”
[218]
Secondly, I observed,
after listening to the experts, that it is easy for them in the heat of the
moment to forget counsel’s often repeated advice. An expert’s role is to
express an unbiased opinion on a technical subject that the Court considers useful
to hear and that is relevant to the subject matter of the case, no more no
less.
[219]
I appreciate the
precarious position in which our legal system places opinion witnesses, and because
they find themselves in such a position, expert evidence can, I believe, be at
times biased. Further, I would add, this consequence is not necessarily the
expert’s fault but is a product of the pressure imposed by the system. Experts
know that it is easier to obtain mandates when their opinions are often accepted
by the courts.
[220]
For example, Glenn R.
Anderson suggests, in his treatise titled Expert Evidence, that not all
expert bias is dishonest or an intentional attempt to mislead or confuse the
court.
Attitudes and expectations inherent in the adversarial system foster
certain beliefs about the role of the expert witness. Some expert witnesses
genuinely view it as their proper role to assist persons employing them by
whatever means is enabled by their specialized knowledge. These experts are
biased, but not necessarily dishonest. They do, however, overlook their primary
duty to assist judges and juries.
[221]
Master Sandler of the
Ontario Superior Court of Justice also discusses this difficulty in Peter
Lombardi Construction Inc. v. Colonnade Investments Inc., a decision in which he gave little weight
to the evidence of two expert witnesses whose objectivity and independence were
compromised due in part to the “exigencies of litigation”.
420 . . . in coming to a decision on any particular
issue where I had conflicting opinions and values from Mr. Doherty and Mr.
Hand, I sometimes accepted Doherty’s views, and sometimes accepted Hand’s
views, and sometimes I accepted neither, and came up with my own assessment
based on the probabilities and reasonableness that surrounded any particular
issue. . . .
[222]
Master Sandler
concluded that the opinion witnesses became advocates for a particular position
on some issues and “let their obligations to their respective clients over-ride
their duty to assist the court in arriving at the correct result”.
[223]
In another case, Master
Sandler gave little weight to the evidence of an advocating expert. He
commented that the expert had gone “too far” in his opinions and took pains to
emphasize that none of the opinion witnesses testifying in that trial were free
of credibility problems such as bias and lack of objectivity. Master
Sandler then cited observations of Feldman J. in Interamerican Transport
Systems Inc. v. Canadian Pacific Express and Transport Ltd.:
61 . . . An expert witness is called to provide
assistance to the court in understanding matters which are beyond the expertise
of the trier of fact. Such a witness is not to be an advocate of one party, but
an independent expert. Expert witnesses are of course paid a fee by the party
calling them which in itself may be considered to affect their independence.
The court will examine the demeanour of an expert in the way the evidence is
given, in particular whether the expert takes on the role of an advocate for
one side, or remains objective, in weighing the evidence and attributing value
to the opinion. If the expert does adopt the attitude of a neutral, then the
fact that he is being paid or that the defendant is his client will cause
little or no concern, but that will not be the case if he appears to lose his
neutrality. In that case the value of his evidence can diminish significantly.
[224]
In Fenwick v.
Parklane Nurseries Limited, MacFarland J. concluded that the expert
witness of the appellant was partisan and gave little weight to his testimony.
35 Courts traditionally afford expert witnesses a great deal
of respect. This is so because these persons possess an expertise in a
particular area of endeavour where lay persons require assistance. The hallmark
of an expert witness is that he or she exercises an independent professional
judgment in their assessment of the facts of a given case. Where there is any
suggestion that a witness who is proffered as an expert has not that
professional independence but has rather simply taken on the cause of the
client who pays the bills, a court will be most reluctant to place great weight
on the opinions of that expert.
[225]
Again, in Shearsmith
v. Houdek, Romilly J. gave little weight to the evidence of an expert who
was more an advocate for a party than an objective opinion witness.
[226]
All in all, judges must
ensure experts are acting in conformity with their role of amicus curiae;
they must decide whether an expert has overstepped and whether the expert’s testimony
is prejudicial to the interest of justice.
As the above examples illustrate, in some cases Canadian courts have simply
disqualified opinion witnesses for playing the role of advocates, and in others,
judges have given little weight to the expert evidence. However, all courts
have held that a lack of objectivity, neutrality and independence has, at the
very least, a significant impact on the weight to be accorded to expert
evidence, if that evidence is not simply declared inadmissible at the outset.
[227]
I believe it is the
duty of a trial judge to be sure that the expert appearing before him has not
advertently or inadvertently put on a counsel’s robe in pressing his opinion upon
the court. Often such improper conduct can only be uncovered through direct
questions from the bench to the expert. Opposing counsel’s cross‑examination
will also often expose such conduct, but not always. In any event, comparing a
judge to a sphinx has been outdated for some time now. Judicial silence is no
longer considered to guarantee impartiality and neutrality in the decision‑making
process. In 1985, Lamer J. (as he then was) observed in Brouillard also known
as Chatel v. The Queen:
17 . . . it is clear that judges are no longer
required to be as passive as they once were; to be what I call sphinx judges. We
now not only accept that a judge may intervene in the adversarial debate, but
also believe that it is sometimes essential for him to do so for justice in
fact to be done. Thus a judge may and sometimes must ask witnesses
questions, interrupt them in their testimony and if necessary call them to
order.
[Emphasis added.]
[228]
This modern trend is
based on the understanding that the principal role of a judge is to discern the
truth. Not all litigants can afford to hire the best counsel or, for that
matter, afford any representation. Accessibility to the courts has been
restricted by costs which have increased substantially. Rinfret J. of the Québec
Court of Queen’s Bench, now the Québec Court of Appeal, described the two
opposing theories as follows:
[UNOFFICIAL TRANSLATION]
The question to be posed is indeed the following: What does “justice”
consist of?
Must a judge, without a word, listen to testimony, hear arguments,
and limit himself to reaching a decision solely based on the evidence and
arguments the parties are willing to submit?
Must a judge, realizing that counsel inadvertently, through lack of
inability, or ignorance, has forgotten to produce evidence or to present an
argument, deliver a judgment he knows to be inequitable to the parties?
Must the client suffer the consequences of the clumsiness of his
counsel?
Some would answer in the affirmative; they believe the judge must
rely strictly and rigorously on what was presented and that counsel, not the
judge, is master of the hearing.
Conversely, the alternate theory demands that the only master of the
hearing be the judge, leaving him to direct the proceedings in the best
interests of justice. To achieve this, the judge must inquire about all facts,
even those which, for one reason or another, a party might have omitted to
submit to the court. He must raise questions of law, even if these have not
been submitted to him, provided that, in each case, he gives the parties or
their counsel the chance to be heard on these issues.
The law or justice is not a matter of surprises or technicalities.
It is a judge’s duty to shed as much light as possible on the
question, to correct the situation, and to make up for the clumsiness or the
ignorance of counsel, should this be required. This is how I understand “justice”.
However, a judge must not act in such a way as to cause the parties
to lose their vested rights, and the judge, by the exercise of his discretion,
will ensure the protection of those rights.
[229]
The judge has liberty
to intervene in the proceedings in the interest of truth, provided he gives both
parties full latitude to address the points raised by his questions. More
recently, the Federal Court of Appeal in NCJ Educational Services Limited v.
Canada (National Revenue) tempered the more traditional view of the role of a judge
expressed in James v. The Queen
by declaring that it does not stand for the principle that a
judge should refrain from intervening, but rather, that a judge should
refrain from excessive intervention. Desjardins J.A. quotes the following
passage from the manual A Book for Judges by the
Honourable J.O. Wilson in support of this more modern view of justice:
. . . the rule is not against any intervention; it is
against excessive intervention. Edmund Burke said: “a judge is not placed in
that high position to be the mere arbiter of parties. He has a further duty,
independent of that, and that duty is to ascertain the truth.
[Emphasis added.]
[230]
I believe that not only
questions for the purpose of clarification are permissible when dealing with
experts, but also questions designed to ensure that the attitude of the expert
witnesses has not become that of an advocate. This broad statement is tempered
by the right that must be fully afforded to counsel to complete the examination
or cross-examination of the witnesses and to answer the questions raised by the
judge.
Framework for the Determination of an Arm’s Length
Price
Defining the Hypothetical Transaction
[231]
The Respondent called
two transfer pricing economists, Dr. Wright and Dr. Becker, as expert
witnesses to opine on the analytical framework for conducting a transfer
pricing analysis. I found each of these witnesses to be unbiased experts. At
times they answered my searching questions in a manner which was not always helpful
to the Respondent’s position. They acted as unbiased experts and not as advocates,
thus avoiding an often critical mistake made by expert witnesses. Both of these
witnesses had in‑depth practical experience in the transfer pricing field
and numerous publications to their credit. I found their opinions to be very
helpful and in conformity with the OECD Guidelines on which the Canadian
transfer pricing rules are based.
[232]
Both of these professionals
agreed that the first step in a transfer pricing dispute is to properly
identify the transactions at issue. In the present case, this step involves
identifying the parties to the controlled transaction, the functions performed
by each party and the risk assumed as part of the transaction.
[233]
Dr. Becker wrote
in his report that “. . . it is necessary to define the terms/characteristics
of a transaction by creating a hypothetical (arm’s length) transaction. This
hypothetical serves as a proxy for the intercompany transaction under arm’s
length circumstances.”
[234]
Dr. Becker
emphasized the importance of matching the risks of the transaction in the
hypothetical construct. Credit risk has a significant bearing on the guarantor’s
potential loss and, as a result, should affect the price charged. This is
different from many other types of related-party transactions where risk may
play a much lesser role. In the instant case, features of the borrowing
transaction, to which the guarantee is accessory, also have a bearing on the
assumed risk.
[235]
Therefore, one of the
first steps of the analysis is the determination of the Appellant’s debt rating
with and without the guarantee. It is expected that the price for the guarantee
would vary with the default risk assumed by the guarantor. The higher the
initial credit rating of the issuer without the guarantee, the lower the price
and vice versa.
[236]
As noted above, the
Appellant did not have a treasury department. GECUS centralized all of the
Appellant’s treasury functions at the head office. Strict instructions were
given by GECUS that only Mr. Werner’s treasury group could raise credit
and deal with outside investors, underwriters and bankers. I infer from the
evidence that this was done because GECUS had to reassure its investors that it
had implemented conservative financial practices throughout its organization.
This behaviour was consistent with GECUS’ AAA credit rating.
[237]
Mr. Werner’s
treasury group determined the timing and negotiated the terms of the
Appellant’s debt issues. As a result, GECUS, the guarantor, exerted control
over the Appellant’s default risk. A third-party guarantor would not control
the financial function of its debtor. It would not determine how and when the
debtor would issue debt. It would not manage the debtor’s cash flow and control
debt payments.
[238]
A third-party guarantor
would not control the timing with respect to its guarantee. The credit
enhancement arrangement would be negotiated in advance for a fixed term and
agreed to by both parties, and the debtor would determine the timing of its
debt issuance. Therefore, a third-party guarantor would assume more risk than GECUS
did in the instant case. For example, commercial paper must be refinanced every
30, 60 or 90 days, depending on the maturity date of the debt. To provide
assurance to the marketplace that the debt can be repaid, investors would need
to know that subsequent debt issues are also covered by the same third-party
guarantee, otherwise the third party could refuse to guarantee new debt issued
to replace maturing debt, or charge a higher premium for the guarantee. As a
standby credit facility would be required to achieve a AAA credit rating, one
can easily imagine that the banking syndicate would require the third-party
credit enhancement to be in place for the full term of the standby facility.
Otherwise, the banking syndicate would be taking on a much greater risk.
[239]
The guarantee
arrangement provided by GECUS to the Appellant was not established in advance
for a fixed duration. The guarantee was given on a one-off basis and only upon
the issuance of the commercial paper, an event that GECUS’ treasury department
controlled to some degree through its performance of the cash management
function for the Appellant. Undoubtedly because the parties were not dealing at
arm’s length, there was no need to agree in advance to a term for the
guarantee. However, this situation could not exist, for the reasons noted above,
if a third‑party arrangement was entered into. Therefore, an arm’s length
guarantor would assume more risk than GECUS, and this would require either an
adjustment in the pricing or the use of an alternative methodology to determine
the arm’s length price. I believe that all of the above factors would be
different for a third‑party insurer or guarantor; they would undoubtedly
magnify the risk for the insurer or guarantor, thus resulting in a higher
premium.
[240]
Mr. Meghji,
counsel for the Appellant, objected that these factors should not be taken into
account in the determination of an arm’s length price on the grounds that GECUS
enjoys control over these factors because of its indirect share ownership of
the Appellant. Simply put, what we have here is a benefit accruing to GECUS qua
shareholder. I believe that Mr. Meghji’s argument is based on an
oversimplification of the corporate law applicable to the Appellant.
[241]
In theory, or in
corporate law, directors manage, or supervise the management of, the business
and affairs of a corporation,
while officers run the daily operations within the framework of the policies
and directions set by the elected board of directors. Yet, in practice, where
economic forces come into play, officers of large corporations determine the
corporate destiny; they have the vision; they hold the reins of the corporation;
and they often select their own successors.
[242]
Shareholders have some
rights; for example, they elect the board of directors. These legal rules can
be altered in a unanimous shareholder agreement and the shareholders can
appropriate the powers of the board of directors “to manage, or supervise the
management of, the business and affairs of the corporation”.
[243]
As for the day-to-day
operations of a business, the Canada Business Corporations Act (the “CBCA”)
does not specifically allow for shareholders to appropriate powers of officers.
According to Bruce Welling, this reality is “formalistically consistent with
the traditional corporate law notion that officers are limited functionaries
appointed by the board of directors”.
[244] In Duha Printers (Western) Ltd. v. Canada,
Iacobucci J. explained:
61 . . . Directors generally owe a
duty not to the shareholders but to the corporation, and shareholders could
not, therefore, control the day-to-day business decisions made by the directors
and their appointed officers. In other words, although the shareholders
could elect the individuals who would make up the board, the board members,
once elected, wielded virtually all the decision-making power, subject to the
ability of the shareholders to remove or fail to re-elect unsatisfactory
directors.
[Emphasis added.]
[245]
All in all, the
fundamental distinction remains that shareholders can appropriate the powers to
appoint the officers, but not the powers of the officers to in fact manage the
business. This is the result of a close reading of subsection 146(1) of the CBCA.
[246]
As the above shows,
GECUS could not appropriate the money management functions of the Appellant by
virtue of the exercise of the voting rights it indirectly held with respect to
the Appellant. Practically speaking, it did appropriate the function, but this
was not something that was attributable to ownership of its shares in the Appellant.
[247]
I also note that due consideration
must be given as well to the fact that the Appellant’s debt had been guaranteed
by GECUS since 1998. The Appellant’s investors had grown accustomed to the fact
that its debt had been guaranteed by its much larger U.S.
parent long before GECUS decided to charge the Appellant a fee for the
guarantee arrangement. In arm’s length negotiations, this fact would not go
unnoticed. A prospective guarantor, when approaching negotiations, would
anticipate that it would be difficult for the debtor to convince its investors
to accept unguaranteed debt on the same terms and conditions as debt guaranteed
by its parent. Investors would attribute less value to the parent’s implicit
support in this scenario; most likely, they would wonder why unguaranteed debt was
now being issued. The cost of borrowing money would likely be higher than it
would be if the Appellant’s debt had never been guaranteed by GECUS. The arm’s
length guarantor could use this knowledge as leverage in negotiating with the
debtor. GECUS and the Appellant are supposed to bargain as arm’s length parties.
This history of the guarantee places the Appellant in a more vulnerable
position, as is shown by the evidence considered later on in these reasons.
[248]
The Crown argues that the
issue herein does not warrant an investigation of the impact of the removal of
the guarantee. According to the Respondent, the trial has to do with the
pricing of the guarantee remaining in place.
[249]
I find this argument ironic
in light of the case presented by the Respondent at trial. It was the
Respondent that insisted that an arm’s length price for the guarantee must be
determined using the yield approach. The Respondent was thereby calling for the
determination of the Appellant’s credit rating both with and without an
explicit guarantee. An accurate determination of the Appellant’s credit rating
cannot be made by ignoring the fact that a guarantee was provided for all of
its prior public indebtedness.
The Crown cannot pick and choose among the economically relevant
characteristics of the transactions and use only those facts that are
favourable to its position. Witnesses for both the Appellant and the Respondent
acknowledged that the prior guarantee is a relevant fact which cannot easily be
explained away in the determination of a credit rating for the Appellant
without an explicit guarantee.
[250]
This brings us to the
next step, which concerns the proper description of the analysis of the parties
to the hypothetical transaction. Dr. Becker suggests the hypothetical
guarantor should have characteristics similar to GECUS, namely, it should be a
AAA-rated multinational subsidiary of a AAA-rated multinational parent.
Similarly, the hypothetical debtor should be a subsidiary of a AAA-rated
multinational corporation. The hypothetical parent of the debtor should be a
corporation conducting an international unregulated financial services business
that borrows large amounts of money in the international commercial paper
markets.
[251]
Dr. Wright emphasized
the importance of evaluating the comparability of the transactions, or the
importance of the methodology used to justify the transfer price, as follows:
A careful transfer pricing study would spend considerable effort to
determine whether GECUS should be compared to a bank or insurance company
guarantor because that affects the guarantee fee computation. The potential
increase in GECUS’ interest expense attributable to default by [the Appellant]
is significantly larger than the guarantee fee that a third party bank or
insurance company would require . . . .
The Correct Methodology
[252]
Both parties
acknowledge there are no comparable uncontrolled transactions. The parties also
agree that the resale price and cost plus methods are inapplicable. Having eliminated
the preferred transaction-based methodologies, each party presented competing pricing
methodologies for my consideration.
[253]
The Appellant introduced
an insurance-based model and a CDS methodology as alternative methods. The
Respondent offered the S&P credit rating methodology for consideration
under the yield approach to establishing the Appellant’s credit rating without
a guarantee, on the theory that the guarantee fee cannot exceed the value of
the benefit resulting from the service provided.
Insurance-Based Methodology
[254]
At trial, following the
examination-in-chief and cross-examination of Mr. Fidelman, I mentioned that
I had considerable reservations about his proposed methodology. First,
Mr. Fidelman admitted that guarantee insurance has been used as a credit
enhancement arrangement only for municipal bonds and asset-backed securities. Such
an arrangement has not been used to mitigate default risk with respect to
corporate bonds or commercial paper. I suspect the market has not developed
because insurers price the risk higher than the benefit perceived by corporate
issuers that have parent corporations able to provide credit enhancement in the
form of a guarantee or other types of explicit support. In the present case, I
believe the risk assumed would not be the same for an insurer as it was for
GECUS, as the insurer would be unable to control the timing, terms and payment
of the Appellant’s debt offering. It would be asked to guarantee GECUS in
advance of its issuance and for a specific term. This was not the case for the
arrangement negotiated between GECUS and the Appellant. In short, the guarantor
would have no control over the debtor’s default risk. This is a situation very
different than that enjoyed by GECUS.
[255]
I do agree with
Mr. Fidelman that, logically, a third-party insurer would attach very
little worth to the “implicit support” of GECUS because if it was called upon
to execute its guarantee of the Appellant’s debt it would be expected to pay
up. Nonetheless, a default by the Appellant on its debt would not be welcomed
by GE and GECUS as it would signify to the marketplace that parts of the
consolidated enterprise were not as reliable as expected. In the end analysis,
GE’s and GECUS’ creditors are lenders on the strength of those corporations’ consolidated
balance sheet. This was confirmed by the Appellant’s experts, Mr. Cole and
Mr. Coombs.
[256]
In addition,
Mr. Fidelman relied upon a credit rating product called “RiskCalc” to
obtain a stand-alone rating. The stand-alone credit rating did not take into account
the implied support of GECUS as it was not designed to deal with debtors in the
financial industry that had AAA-rated parents. Mr. Fidelman was required to do an arbitrary
“notching up” to match the public ratings of independent companies, which calls
into question the reliability of this methodology. Finally, Mr. Fidelman
concluded that the guarantee fee under the insurance approach would still be
0.85% even if the Appellant was viewed as a AAA-rated issuer without an
explicit guarantee. This stands in contradiction to the position adopted by Mr. Werner,
who believed an arm’s length guarantee fee could not exceed the value of the
benefit received by the Appellant.
[257]
It is impossible for me to
determine with any degree of certainty whether the above factors, some of which
support the insurance methodology and others of which clearly do not, are
perfect counterweights to each other. Mr. Fidelman did not address how the
additional risk that would be assumed by the insurer would affect price. I
surmise that if he did not address the fact that a third-party insurer would
assume greater risk under a properly structured arm’s length guarantee
arrangement, it was on the advice of counsel, given the legal argument that I had
rejected earlier. Therefore, I consider the insurance-based methodology to be
unreliable in the circumstances, save perhaps for its use as one method among
others to be considered at the stage of the “sanity check” recommended by
Dr. Wright as a final step to be performed in ascertaining the arm’s
length price for the transaction.
CDS Methodology
[258] Dr. Hull uses a CDS methodology to
come up with a price range for the guarantee transaction. A careful review of
Dr. Hull’s testimony reveals that the CDS approach is analogous to the
yield approach proposed by the Respondent. This witness admitted that he was
asked to use for the purpose of his analysis an assumed credit rating which was
provided to him by counsel for the Respondent. On the basis of the assumed
rating, he concludes that the premium under a CDS would be equal to the yield
difference between the Appellant’s assumed rating and the AAA rating attributed
to the Appellant’s debt as a result of GECUS’ guarantee. In the end, Dr. Hull’s
analysis does little to clarify the issue herein. The accuracy of his
conclusion is entirely dependent on the accuracy of the assumed credit rating
that he uses in his report. That is the crux of the matter in this case.
Yield Approach
[259] The Respondent asserts that the first step
in the transfer pricing inquiry is to calculate the value of the benefit
enjoyed by the Appellant as a result of GECUS’ guarantee. This should be done
using the yield approach. The benefit is equal to the interest cost savings for
the Appellant determined by comparing the interest cost of unguaranteed debt to
that for guaranteed debt. To determine the interest savings for the Appellant,
one must arrive at a factual finding of the Appellant’s credit rating without
the explicit support of its parent. Could the Appellant be a AAA-rated issuer
without the guarantee?
[260]
Counsel for the Appellant
takes issue with this approach on the grounds that it does not entail a search
for an arm’s length price. The Appellant argues that this method leads to the
determination of a value to the owner. In addition, counsel argues that the
Crown has priced the wrong transaction. According to counsel, for the purpose
of the appeal the Minister is required to price the actual transaction, which
is a guarantee, and not a loan as suggested by the Respondent.
[261]
This latter argument is
contradicted by the testimony of the Appellant’s principal lay witness, Mr. Werner,
who testified that he determined the guarantee fee on the basis of his estimate
of the benefit received by the Appellant:
. . .
Mr. Meghji: Now, Mr. Werner, what did you understand your objective to be in
determining the guarantee fee?
Mr. Werner: I understood my objective to be to develop and support a guarantee
fee which was arm’s-length, which was a market rate, which was supportable by
data, objective data, in the marketplace.
Mr. Meghji: Now, in seeking to arrive at this, you have used the words market
price. I will use the words market price.
In arriving at this market price of the guarantee, the guarantee fee
amount ‑‑ sir, I want you to go in bite-sized pieces carefully --
what exactly did you do, you and your staff?
Mr. Werner: We started with -- we started by determining what the benefit
to the borrower, that is GE Capital Canada, of the guarantee would be.
Mr. Meghji: Okay. And in the process of starting with that, did you have to
make any -- did you have to make any analysis of whether GE Capital Canada was
an investment grade company?
Mr. Werner: We talked about this at the end of -- the answer is yes. We talked
about it at the end of the day yesterday that we had determined in 1988, 1989
that GE Capital Canada was, at best, right on the edge between investment grade
and non-investment grade, and that our opinion as to that creditworthiness had
not changed.
Mr. Meghji: So you started with that, with that this was not an investment
grade company?
Mr. Werner: Yes.
Mr. Meghji: Then you said you started by -- the question was asked: What was
the benefit to GE Capital Canada?
Mr. Werner: Yes.
Mr. Meghji: Now, how is it that you went about determining the benefit to GE
Capital Canada of the guarantees?
Mr. Werner: Well, they certainly could not -- would not, in a commercial
setting, pay more the guarantee fee than the benefit of that guarantee
fee.
. . .
[Emphasis added.]
[262]
In his testimony, Mr. Werner
has admitted that the yield approach proposed by the Crown is in fact an
acceptable methodology for determining an arm’s length price. As his own words
show, that is what he used himself.
Application of Credit Rating Methodology
[263]
The Crown posits that the
Appellant did not benefit form GECUS’ guarantee of its indebtedness because its
debt would have been rated AAA solely on the basis of the implicit support of
GECUS. The Respondent had Mr. Emmer qualified as an expert on credit
rating methodology and the Appellant had Dr. Chambers similarly qualified.
Both of these experts held senior positions with S&P over the course of
their long careers with that credit rating agency.
[264]
Both of the experts
relied to various degrees on the rating criteria published by S&P and
Moody’s to demystify the rating process. The contrary opinions expressed by the
experts are largely based on two factors. Dr. Chambers prefers the
relevant quantitative factors over the applicable qualitative factors. Dr. Chambers,
who appeared for the Appellant, places greater weight on factors such as the
small size of the Appellant’s balance sheet compared to that of its parent and
its high debt-to-equity ratio of 12 to 1 in concluding that the
Appellant did not possess the characteristics to be rated AAA. It was on
the cusp of an investment-grade rating once account was taken of a certain degree
of implicit support from its parent, GECUS.
[265]
Mr. Emmer arrives
at the contrary view because he favours qualitative over quantitative factors
in his analysis. In the present case, Mr. Emmer suggests that GECUS would
have suffered catastrophic damage to its reputation had the Appellant been
allowed to default on unguaranteed debt it issued to the public. According to
this expert, GECUS would have taken whatever steps were necessary to ensure
that no default would occur because, if it failed to do so, its own prized AAA
rating would have been jeopardized. The cost to it of a credit rating downgrade
would far exceed the cost of preventing the Appellant’s default. GECUS gave its
guarantee out of convenience, for itself alone. The guarantee facilitated the
task of its treasury department with respect to debt issuances by the
Appellant.
[266]
I have difficulty with
Mr. Emmer’s opinion on this matter. I do not believe the quantitative
factors identified by Dr. Chambers in his report can be summarily
dismissed on the theory of reputational pressure. The evidence shows that
S&P, the rating agency where Mr. Emmer spent his career, demanded that
GE execute a keep-well agreement under which it committed to maintaining GECUS’
debt‑to‑equity ratio at 8 to 1. GE and GECUS both enjoyed
AAA ratings in the taxation years under review. I understand that the rationale
for the keep-well agreement was tied to the level of preferred share capital
that GECUS had issued to outside investors. Preferred shares have debt-like
characteristics. They are not considered to offer the same type of permanent
capital as common shares. Events can trigger the redemption of the shares
unless they have been made to be perpetual. To protect against an erosion of
capital, the keep-well agreement required GE to, inter alia, add capital
if preferred shares were retracted and not replaced. I imagine
that, if GECUS’ debt-to-equity ratio was unimportant, S&P would not have
seen the need to demand the execution of the keep-well agreement.
[267]
Mr. Emmer and the
Crown dismiss the debt-to-equity ratio on the basis that it is circular in its application.
In other words, GECUS can make the Appellant’s balance sheet look more or less
attractive by providing it with more or less capital. GECUS did not provide
more capital to the Appellant because the guarantee saved it from doing so.
Whether this is true or not does not matter. I believe creditors and rating
agencies are concerned about the debt-to-equity ratio of debtors. Common share
capital, in my opinion, provides a more robust means of weathering either a
storm of financial difficulty that affects the industry or a generalized
recession than does implied support. Creditors have no recourse against the
parent if they rely on the expectation that the parent will come to the rescue
of its subsidiary and it fails to do so. I infer that S&P acknowledged this
fact when it demanded that GE execute a keep‑well agreement with respect
to its direct subsidiary, GECUS.
[268]
I also take issue with
the suggestion that GECUS could have injected capital into the Appellant to improve
the latter’s debt-to-equity ratio; I do so on the grounds that this contradicts
the well-accepted principle that a corporation is a separate person whose very
existence provides limited liability protection to its shareholders. The extent
of a shareholder’s exposure through the corporation is limited to the amount of
capital the shareholder chooses to invest. In the absence of a guarantee from
the shareholder, creditors can expect nothing more or less.
[269]
This comment must be
tempered with respect to other types of corporate entities that exist on the
Canadian corporate landscape. For example, shareholders of an Alberta unlimited liability corporation (“ABULC”) have
unlimited joint and several liability for any liability, act or default of the
ABULC. This type of entity is often used in cross-border financing with the U.S. to facilitate so-called “double dip” financing
structures.
[270]
Because the shareholder
is jointly liable with the ABULC, a guarantee of the corporation’s debt by the
shareholder might not create a benefit for the corporation. Moreover, because
the joint and several liability results from share ownership, there might be no
service that the shareholder could charge for; it would be passive association
as a shareholder. If the shareholder had a higher credit rating than the
subsidiary, legal status alone may require a notching up in the credit rating
of the subsidiary. This may also have implications for the interest rate that
could be charged on intercompany loans. This latter comment must be qualified
if the shareholder is simply an intermediary shell corporation whose only asset
is shares of the ABULC. This might also be true if an indirect parent of an
ABULC provides a guarantee for its indebtedness, assuming, for example, the
direct shareholder is a non-resident corporation. The guarantee in these
circumstances might benefit both the ABULC and its shareholder, which might
necessitate allocating the liability equally.
[271]
The immediately
preceding comment must be further qualified with respect to a Nova Scotia unlimited liability company (“NSULC)”, which is often
used in a cross-border financing context for the purpose mentioned above. In
this case, the liability of the shareholder is conditional.
[272]
Briefly, the Nova Scotia Companies Act (“NSCA”)
permits the incorporation of a company with no limit on the liability of its
shareholders. Pursuant to section 135 of the NSCA, past and present
shareholders of a NSULC are jointly and severally liable for all debts and
liability when the NSULC is wound up or liquidated with insufficient assets to
satisfy its obligations. For past shareholders, the liability is extinguished
if the shareholder ceased being a shareholder one year before the winding-up or
liquidation.
[273]
An NSULC must be wound
up or liquidated if it has insufficient assets to meet its obligations. There
is no immediate liability for the shareholder. In this context, the guarantee
might provide a greater benefit to the corporation and a more remote benefit to
the shareholder. This would be particularly true if the shareholder was a shell
corporation. Needless to say, the facts and circumstances of each case must be
well understood. With this background in mind, one must be aware that it would
be dangerous for taxpayers to draw general inferences from this particular case,
as differences in facts or circumstances or in the economically relevant
characteristics of a transaction can lead to a very different result. In the final
analysis, transfer pricing is largely a question of facts and circumstances
coupled with a high dose of common sense.
[274]
Mr. Emmer also
used the argument of reputational pressure to skip over an essential first step
imposed under the rating methodology criteria used for rating subsidiaries of a
public corporation. The rating literature submitted at trial provides that an
analyst must start the rating process by completing a stand-alone rating of the
subsidiary before proceeding to determine whether ratings uplift is required in
order to take into account the benefit of implicit support by the parent. The
literature explains the rationale for the stand‑alone rating; its purpose
is to allow the analyst to gauge the rating gap between the subsidiary’s and
its parent’s ratings. More concrete signs of implicit support are required to bridge
a large rating gap. An analyst cannot make a rating proposal to the ratings
committee that oversees the final rating until this step is completed.
Mr. Emmer knew this requirement, yet he does not fully comply with it in
his main report. I find nothing in the literature that would require this step
to be carried out only with respect to certain types of industries. Because he
fails to conduct this critical first step, he has no reference point for
opining on the rating gap and the signs of implicit support that must be
recognized by the analyst to justify the notching up of the Appellant’s rating.
With this background in mind, his opinion on the subject appears to me to be
speculative at best.
[275]
Dr. Chambers, on
the other hand, did perform a stand-alone rating analysis as prescribed by the
S&P criteria in effect during the years under review. He concludes, after
applying all of the relevant criteria, that the Appellant would have been rated
B+ to BB- on a stand-alone basis. As a result, he is in a position to determine
that there is a large differential between the Appellant’s stand-alone rating
and that of its parent. The gap is between 12 and 13 notches. He testified
that the typical ratings uplift is two to three notches.
[276]
There are other
difficulties with the influence that Mr. Emmer claims reputational
pressure would have had on GECUS’ behaviour. I view reputational pressure in
this context as an offshoot of the buildup of social capital in the marketplace
by GE and GECUS. All parties agree that AAA‑rated issuers must
demonstrate consistent behaviour to achieve the confidence of their debt holders
that all of their obligations will be promptly discharged. There is
disagreement among the parties as to the extent to which this consistent
behaviour must be displayed, the Respondent’s position being that it must be
extended to the Appellant because it is a core or strategically important subsidiary.
[277]
On cross-examination,
Mr. Emmer was confronted with the S&P rating report issued in 1999 for
the GE group. GE Financial Corp., an indirect insurance subsidiary of two AAA-rated
parents, GECUS and GE, was rated A+, that is, four notches below the AAA rating
of its parent corporations.
I note that GE Financial Corp. had 10 times more assets than the
Appellant; its assets represented approximately 21% of GECUS’ consolidated
assets. If reputation alone is a strong factor for ratings equalization, surely
GE Financial Corp. would have been a much better candidate for the ratings
uplift proposed by Mr. Emmer for the Appellant. A default by GE Financial
Corp. would no doubt have had a much bigger impact on GECUS than a default by
the Appellant. I recognize that GE Financial Corp. is a regulated insurance
company and that the Appellant operates an unregulated financial services business.
Nonetheless, I cannot imagine how GE Financial Corp.’s failure would be of lesser
consequence to GECUS than a default by the Appellant. I share
Dr. Chambers’ view that the sheer size difference alone would account for
a bigger impact.
[278]
In addition,
Mr. Emmer made a computational error in his application of a quantitative
fact. Mr. Emmer admitted Dr. Chambers had correctly pointed out that
he failed to take into account currency differences when comparing certain of
the Appellant’s financial ratios to those of its parent. When corrected for
these errors, the Appellant’s growth ratio and the size of its balance sheet
compared to GECUS’ consolidated balance sheet were revised downwards.
[279]
I believe Mr. Emmer
also failed to adequately consider the impact of the removal of GECUS’
guarantee. This guarantee had been in place long before there was a hint of a
guarantee fee. There is no longer any suggestion by the Crown that the
guarantee was executed as an artifice for the siphoning off of the Appellant’s
profits in the form of a guarantee fee. It was executed solely for legitimate
business reasons.
[280]
Dr. Chambers
believes that the investment community and the rating agencies would have
reacted negatively to the removal of the guarantee. I agree that, if GECUS had removed
its guarantee, that would have signified that its appetite for providing credit
support for the Appellant had diminished. I imagine that it would have been
difficult in those circumstances to convince investors that they did not have
to worry about the removal of the guarantee because reputational pressures
would compel GECUS to avoid the Appellant’s default in any event. Moreover, I
believe the investment community would have reacted by quoting the well-known
proverb “a bird in the hand is worth two in the bush”.
[281]
Implicit support is
nothing more than one’s expectation as to how someone will behave in the future
because economic reasons will cause the person to act in a certain manner.
Economic circumstances can change quickly, as evidenced by the recent credit market
meltdown. A guarantee is a much more effective form of protection. It is
something that investors in the present case would have been reluctant to give
up in light of the fact that substantially all of the Appellant’s debt had been
guaranteed for a very long period of time.
[282]
Mr. Emmer,
appearing for the Respondent, recognized that he had difficulty imagining how a
decision to drop the guarantee could have been presented as a positive or
neutral fact by the Appellant in attempting to secure commitments for
unguaranteed commercial paper. For example, in response to my question on this
point, he testified as follows:
JUSTICE HOGAN: The facts are a little different. I think counsel is
putting the question to you that they come to you and you are wearing your
cynical, sceptical hat, and they say: We have been guaranteeing our debt for
eight years now and we want to remove the guarantee. Can we get there?
Wouldn’t your antennas go up a little bit, and say, why --
wouldn’t you ask these very questions, Why won’t you put pen to paper? Why are
you sort of changing your attitude?
THE WITNESS: . . . It would have to ask: Why would
they -- other than for this particular instance, why would they want to stop
guaranteeing the debt of their subsidiary?
I understand why we’re looking at it that way, and I
think sort of my thought process was, and still would be, that if they allowed
them to issue the debt, that they would stand behind it with a guarantee or
without a guarantee.
However, it is difficult to come up with a scenario
where they would not --where they would cease to guarantee the debt, other than
for purposes of the discussions that we have been having for the past three or
four weeks.
So I have problems sort of making that leap, because I
agree it is an unrealistic scenario. It is very hypothetical.
Why would they want to stop? You know, you look at the
infrastructure that they have in place up in Stamford,
Connecticut, the trading desk and all of the
relationships that they have with commercial paper dealers, and so on. And why
would they want to stop doing it?
So I think it is an unrealistic scenario that you are
asking me to really opine on. So that is why I keep coming back to the same
point, that if they let them do it, if they let -- sorry, if they let Canada
issue without the guarantee, they would stand behind it.
The question is: Why would they want to do that? And
it is very difficult for me to come up with a reason why -- it would be an
irrational decision on their part to do that.
So that is the challenge that I guess we’ve all been
having for the past couple of weeks. It becomes very, you know, hypothetical
and, you know, a bit unrealistic to assume that.
I mean, everything -- as I understand it, everything
that GECC does is -- you know, I think they have, I believe we discussed, Australia and New Zealand, and I forget how
Japan was structured, but they
guaranteed that.
My antenna would go up, also. Before mine would go
up, I know that Mr. Werner’s would be up well before mine, and he would
say, Is this a reasonable thing for us to do? And I think he would conclude it
is not a reasonable decision; therefore, he wouldn’t --
JUSTICE HOGAN: He’d say, Why rock the boat; right?
THE WITNESS: Exactly, yes, yes. Why (a) rock the boat, and (b) I
think, you know, even if we assigned an AAA rating to the debt, I still
think that investors would want a premium.
JUSTICE HOGAN: They would want a premium?
THE WITNESS: They would want a premium without the guarantee, yes. I
mean, we have been here for a month or three weeks -- it is seems like a month
– like, three weeks discussing this, and we still haven’t come to a definitive
conclusion.
We could rate it AAA, but that doesn’t mean that
investors wouldn’t want some kind of premium. The guarantee is clear. It is
black and white. Here it is. You don’t have to go through 15 criteria and
sort of count angels on the head of a pin to see if it is core, strategic,
non-strategic, or what have you.
[Emphasis added.]
[283]
During this line of
questioning, I noted from Mr. Emmer’s demeanour that he was very
uncomfortable answering the question. For example, he claims that the removal
of the guarantee is an unrealistic scenario. For the reasons that I noted earlier,
this question lies at the very foundation of the Crown’s case submitted at
trial. It is an economically relevant fact that should have been considered by
Mr. Emmer in preparing his opinion.
Was the Guarantee Necessary?
[284] The last point on this subject is whether
market participants would have purchased the Appellant’s unguaranteed debt even
if it was rated AAA. Generally speaking, I could address this point simply by
referring to my earlier conclusion that the evidence does not show on a balance
of probabilities that the unguaranteed debt of the Appellant would be rated
close to AAA. Nonetheless, I will deal with this specific matter because this
case is the first of its kind to be presented to the Court for consideration
and there is a substantial amount of money at stake.
[285]
Mr. Coombs
testified that the banks take third-party ratings into account. However, he did
note that these ratings are only one of the many inputs considered in the banks’
internal credit assessment process. He also emphasized that the banks’ internal
rating will often be lower than the ratings of the external rating agencies.
[286]
I do not find it surprising
that banks adopt a more cautious approach. They are committing capital while
the rating agencies are being paid to provide a credit rating opinion by the
debt issuer that they are asked to rate. This raises numerous issues of
conflict of interest which are beyond the realm of this judgment. Suffice it to
say that recent events such as the sudden meltdown of asset-backed financing structures
have shown that these concerns are nonetheless very legitimate.
[287]
Mr. Lewis noted in
his testimony that sophisticated market participants such as Texaco would not
assign weight to implicit support because they would be accepting a risk
without a commensurate return. Security of principal repayment was paramount
and implicit support was simply an extrapolation of someone’s opinion that
economic incentives would cause the parent company to act although not legally
bound to do so. Implicit support is like a metaphorical “invincible wallet”. It
is something investors believe exists and may be available to provide financial
support if the right circumstances are present, but few investors are foolish
enough to believe that it is equivalent to a guarantee. Mr. Emmer
recognizes this very fact in his testimony cited in paragraph 282 above. By its
nature, implicit support does not afford to sophisticated investors the same degree
of reassurance that the parent will act that a legally enforceable guarantee
provides.
[288]
I note that the
commercial paper market is largely composed of sophisticated investors, whether
they be corporations that have a need to invest excess short-term liquidities,
large pension funds or conduit entities such as mutual fund trusts that employ
experienced money managers. Retail investors often participate in this market by
buying units of a mutual fund trust thereby benefiting from the market know-how
of the funds’ money managers. In light of this evidence, it appears implausible
that the Appellant could have raised the same large sums of money at the low
interest rates that it benefited from even if its debt had been rated AAA.
Mr. Emmer confirmed this fact during his cross‑examination. Dr. Booth
also confirmed that most of the Canadian subsidiaries of foreign public
corporations borrow in the Canadian commercial paper market on the strength of
a parent company guarantee. This fact was undisputed by the Crown. Therefore,
it is hard to imagine there would be the same willingness for investors to
accept AAA-rated debt pricing for unguaranteed debt of the Appellant.
[289]
Dr. Saunders and
Mr. Meyerman also testified, inter alia, that economic incentive
was the most important factor in determining the likelihood of parental
support. These witnesses confirmed the fact that GE and GECUS valued their AAA
status. This rating was the key element of both companies’ competitive
strategies, but more so for GECUS than GE because the ability to offer low
interest rates is an important competitive advantage for leasing and other
forms of equipment financing.
[290]
I have no doubt that GE
and GECUS valued their AAA status. Mr. Werner confirmed this fact in his
testimony. This being said, I believe, for the reasons noted above, that it
would be an unwarranted leap of faith to conclude that the Appellant’s credit
rating would be equalized with that of its parent if there were no guarantee in
place. I note that neither Dr. Saunders nor Mr. Meyerman is an expert
in credit rating agency methodologies.
[291]
As stated earlier, the
Crown’s theory that the guarantee was unnecessary contradicts Mr. Werner’s
business judgment that the guarantee was necessary, a position acknowledged to
have merit by Mr. Emmer during his cross-examination. In other areas involving
the application of the ITA, the courts have declared that they will
exercise caution before adopting conclusions that contradict the properly
exercised business judgment of taxpayers. This approach is preferred in order to
avoid the proverbial “Monday-morning quarterback syndrome”. In the present case,
the experts were asked to provide an opinion after the fact based on incomplete
information. They were not in the business trenches. It is hard to establish
the exact dynamics of credit markets after the fact. They change quickly.
[292] In making these comments, I note that the
wording of section 247 of the ITA appears problematic when attempting to
apply the notion of business judgment. The problem arises from the fact that
subsection 247(2) allows the Minister to scrutinize certain cross-border
transactions and essentially substitute an arm’s length price for the price
used in the transaction. I make reference to the business judgment rule merely
because it is a useful tool for evaluating the credibility of a witness. Other
corroborating factors, such as Mr. Emmer’s recognition of the legitimacy
of the guarantee, assist me in according greater evidentiary weight to
Mr. Werner’s testimony. Finally, as a passing comment, I note that
Mr. Werner has been retired from GE and GECUS for a number of years. The
passage of time has freed him from the influence of his former employer and
undoubtedly has contributed to his objectivity on the subject of the guarantee.
[293]
Mr. Werner was busy
raising billions of dollars of funds for the Appellant when the issue of the
guarantee was being considered. He dealt with the major financial market
intermediaries on a daily basis. Mr. Meyerman acknowledges that Mr. Werner
was very good at his job and that GECUS treasury personnel was often more
experienced and qualified in the workings of debt markets than the personnel
working for the major banks.
Mr. Werner was being judged on the results he achieved in a corporate
culture that punished failure quickly and rewarded long-term success. For this
reason, greater weight must be accorded to the testimony of Mr. Werner on
this point.
[294]
Mr. Werner’s view
is supported by the testimony of Mr. Coombs. Mr. Coombs, given his Canadian
banking experience, was much more familiar with the workings of the Canadian
capital markets than Mr. Meyerman, who worked predominantly in the U.S. Recent events illustrate that there are differences
in capital market practice in different countries. Canadian banks are more
conservative in their lending practices than U.S.
banks, as confirmed by Mr. Coombs in his testimony.
[295]
Mr. Meyerman
accepts this view although he qualified his opinion by saying it is only the TD
Bank that is known to be conservative. There is no debate on the issue that
standby bank facilities must be in place to support debt issued in the
commercial paper markets. The facilities are required to guard against systemic
market risk, including liquidity risk. Investors in commercial paper need reassurance
that on the maturity of their paper the issuer will be able to repay. For
example, if an issuer was unable to issue new commercial paper because of a
flight to quality, the standby facilities must be available to be drawn upon to
repay the maturing commercial paper. This reassures investors that there is
security of principal repayment notwithstanding the possibility of short-term
market disruptions.
[296]
Mr. Coombs was
adamant in his testimony that the Appellant could not arrange standby facilities
of the size required to cover the large volume of debt that it needed to issue in
order to execute its business plan. Mr. Meyerman acknowledges Canadian
banks would have had to play a large role in this type of credit facility. I
cannot imagine how a large banking syndicate could have been put together to
support a facility sufficient to cover the commercial paper issued by the
Appellant, in light of the fact that the Appellant’s debt had been guaranteed
by GECUS for many years. If the answer is that economic incentives would cause
GECUS to support the Appellant in any event, I can easily imagine that, as
suggested by Mr. Coombs, the banks would say “put your money where your mouth
is”.
[297]
The Crown, in argument,
attempted to minimize this point by asserting that GECUS had standby facilities
in place that the Appellant could rely on. I cannot accept this point of view.
GECUS, as a separate legal entity, is not required to give the Appellant access
to its credit facilities. If it did so through contractual agreement, it would
have had to negotiate this right with its banking syndicate. Undoubtedly, GECUS
would have been expected to remain liable for amounts drawn by its subsidiary,
the Appellant. This would have become a different form of explicit support that
would have constituted a bona fide group service that GECUS would have been
entitled to charge an arm’s length fee for.
[298]
Dr. Saunders was
the only expert witness for the Respondent who provided evidence as to the
market price of the guarantee. His opinion on the guarantee fee was premised on
his finding that the Appellant’s final credit rating would have been AA+
without an explicit guarantee. I do not accept his conclusions on the
Appellant’s final credit rating for many of the same reasons that I do not
accept Mr. Emmer’s view on this subject. Briefly, Dr. Saunders failed
to consider the impact of the removal of the guarantee. He did not perform a
stand-alone rating in his initial report. His conclusion is at odds with the
business judgment of Mr. Werner. In the case of this witness I would add
that he acknowledged he was not an expert in applying credit rating
methodology. In short, I prefer Dr. Chambers’ and Mr. Werner’s
evidence on this point.
[299] The methodology used by Dr. Saunders
to value the guarantee fee also appears to be flawed. He used historical
default data from rating agencies for a AA credit rating to arrive at the
expected loss reserve. He argues that the Appellant’s score would have been less
than 100 basis points based on his opinion of its credit rating, which I
have found to be flawed. Assuming nonetheless that he is right in his
assumption, he fails to take into account a rate of return on the risk capital,
which is required as compensation for expected loss. Because Dr. Saunders
allows a charge only for expected loss, he has not made allowance for a return
in the form of profit. This methodology allows the guarantor to recover only
its cost. In the arm’s length world, a service is generally provided for
profit. A guarantor would not risk capital with the hope of only breaking even.
If it did, it would be better off investing its capital in risk-free term
deposits.
[300]
Dr. Saunders’
analysis also incorporates the wrong risk capital requirement. He uses the risk
capital requirement of the Basel II regime and not that of the Basel I regime which was in force in the years in
question. When the Basel I requirement is applied in conjunction with Dr. Saunders’
opinion that the Appellant would be rated AA, which I have rejected for the
reasons outlined earlier, GECUS’ expected return on capital is lowered from
149.3% to 12.5%. When adjustments are made to Dr. Saunders’ “capital at
risk” model to take into account a BB+ final rating, the expected return on
capital at risk is reduced to 7.5%.
[301]
For all the reasons
noted above, I conclude that the Appellant’s final credit rating without
explicit support would be in the range of BBB‑/BB+. This means
the ratings uplift from the stand‑alone/status quo rating to take
account of implicit support is three notches.
Are Further Adjustments Required to the Yield
Approach?
[302]
The Respondent argues
that the determination of the yield spread is only the first step or starting
point for the determination of the value of GECUS’ explicit guarantee.
According to the Respondent, further adjustments are required to take into account
the benefits that the guarantee provides for GECUS. The Crown argues that the
yield approach creates a “value to owner” benchmark. This is an inappropriate
test for determining an arm’s length fee. In its written submission filed at
trial, the Respondent identifies the following points as potential benefits
that can be raised by the Appellant, in its hypothetical arm’s length
negotiations with GECUS, for the purpose of lowering the guarantee fee that it
should pay:
a.
no management of the risk;
b.
it is in [GECUS’] best interest to support the
Appellant, even in the absence of an explicit guarantee, because of the
potential costs involved in failing to support, including loss of AAA leading
to higher interest costs, loss of clients and inability to roll over commercial
paper;
c.
[GECUS’] easier access to, and freedom with
respect to, the Appellant’s capital since there are no restrictions arising as
there might be with other forms of guarantee;
d.
the Appellant is part of the hub and spoke
strategy favoured by [GE]; a weakening of a spoke would weaken the whole wheel;
e.
no cost to [GECUS]; and
f.
if a third party provides a guarantee, [GECUS]
will be relieved of its obligation and should therefore contribute to paying
the cost of that third-party guarantee.
[303]
The second, third and fifth
of the above points made by the Crown are all covered under the yield approach
because the final rating reflects a three‑notch ratings uplift. In short,
the reputational pressures and economic incentives that motivate GECUS to
provide financial support are already fully taken into account in the yield
approach analysis. A second adjustment would be tantamount to double counting
those factors.
[304]
The first point has
been dealt with in my analysis of the insurance methodology proposed by
Mr. Fidelman. I believe a third-party insurer would be unwilling to
evaluate the risk at the same price as GECUS did because it would be assuming
greater risk. I also subscribe to the view that a third-party insurer would
place less weight on implicit support because the fact that a third-party
guarantee was obtained would be a strong indication that the parent would be
unwilling to support its subsidiary in the event of failure or of default on
the debt instruments covered by the third-party guarantee. The insurer would be
expected to pay up in such a case because it received a premium for the service
of having taken on the risk of the Appellant’s default. The two factors
commented on immediately above in this paragraph offset each other to some
degree. This being said, it is too difficult to determine whether they are
perfect counterweights to each other. As noted earlier in this judgment, that
is why I chose the yield approach over the insurance methodology.
[305]
I agree with
Mr. Werner’s view that the Appellant cannot be expected to pay 100% of its
interest cost savings. Otherwise, it would have no economic incentive to enter
into the transaction. With this in mind, I note that, under the yield approach,
the interest cost savings based on the rating differential between BBB‑/BBB+
and AAA, the latter being the rate achieved with the GECUS guarantee in place,
work out to approximately 183 basis points or 1.83%. I am of the
view that a 1% guarantee fee is equal to or below an arm’s length price in the
circumstances, as the Appellant received a significant net economic benefit
from the transaction. The net economic benefit exceeds the 1.83% calculated under
the yield approach.
Without a guarantee, the Appellant would have been unable to procure standby
letters of credit in an amount sufficient to cover its commercial paper
program. It is undisputed that the Appellant did not reimburse GECUS for the
costs incurred by the latter under its standby facility. The Appellant would
have been unable, in the absence of the guarantee, to execute its business plan,
as the Canadian commercial paper market was geared to the highest investment-grade
issuers.
Part XIII Withholding Tax
[306]
The Crown asserts that
the Minister issued the Part XIII assessments on the grounds that the guarantee
fees are benefits that the Appellant conferred on a shareholder. These amounts
are deemed to be dividends for the purposes of Part XIII.
[307]
This position is not
sustainable in light of my conclusion that the guarantee fee paid by the
Appellant did not exceed the amount of an arm’s length price. It is undisputed
that the Appellant paid the full amount of withholding tax due on the guarantee
fee on the basis that the amounts were deemed to be a payment of interest for
the purpose of Part XIII. Therefore, there are no grounds for upholding the
Part XIII assessments.
CONCLUSION
[308] For all the reasons set out above, I allow
the Appellant’s appeals and I hereby order that the assessments under Parts I
and XIII of the ITA covered by these appeals be vacated.
[309]
The
cases cited and considered by
the Court are listed in the appendix hereto.
Signed at Montréal, Québec, this 4th day of December
2009.
"Robert J. Hogan"
Email this Content